ZacharyZhang-NY/Kigi-CLI
1
0
Fork 0
Code Issues Pull Requests Actions 13 Packages Projects Releases Wiki Activity
Files
553ec1c62096b7fd0a78940595b67bcacac94ef2
Kigi-CLI/crates/codegen/kigi-models/src/lib.rs
T
ZacharyZhang-NY 649c5f5641 fix(models): drop the ultra reasoning effort, which no backend accepts 
The hardcoded Codex catalog advertised `ultra` for gpt-5.6-sol and
gpt-5.6-terra, so selecting it sent `reasoning.effort = "ultra"` and the
Responses endpoint answered 400 with its real menu, which ends at `max`.

The tier was kigi's invention: the enum documented it as codex-only, the
catalog was its only producer, and the upstream model snapshot never
listed it. Removing the variant as well as the catalog entries means
`/effort ultra` is rejected at the TUI instead of parsing and failing on
the wire; a persisted `ultra` degrades to the model default through
`lenient_reasoning_effort_opt`, as designed for vocabulary changes.
2026-07-27 12:12:14 -04:00

3067 lines
124 KiB
Rust
Raw Blame History

//! Kimi model catalog primitives (PRD F2/F4).
//!
//! This crate owns:
//! - the compiled-in platform registry ([`PlatformId`] + its spec rows): the
//! Kimi Code subscription channel plus the API-key platforms;
//! - the `GET {base}/models` wire contract ([`WireModel`]) and the capability
//! derivation ported from kimi-cli `auth/platforms.py`;
//! - the managed catalog key format `{platform_id}/{model_id}`;
//! - the bundled OFFLINE-LAST-RESORT fallback catalog
//! (`default_models.json`), used only when the live `/models` sync fails
//! AND no disk cache is usable. Every id in that file is sourced from
//! kimi-cli 1.49.0 (see the module docs on [`DEFAULT_MODELS_JSON`]).
//!
//! At runtime each model is resolved via:
//! CLI flag > ENV var > config.toml > server-delivered > these defaults
use std::sync::LazyLock;
pub mod enrichment;
// Platform registry (PRD F2)
/// Env var holding the moonshot-cn API key (wins over the generic name).
pub const MOONSHOT_CN_API_KEY_ENV: &str = "KIGI_MOONSHOT_CN_API_KEY";
/// Env var holding the moonshot-ai API key (wins over the generic name).
pub const MOONSHOT_AI_API_KEY_ENV: &str = "KIGI_MOONSHOT_AI_API_KEY";
/// Generic moonshot API key env var, applied to BOTH open platforms when the
/// platform-scoped name is unset.
pub const MOONSHOT_API_KEY_ENV: &str = "KIGI_MOONSHOT_API_KEY";
/// Base-URL override for moonshot-cn (dev/test escape hatch mirroring
/// `KIGI_CODE_BASE_URL`; production uses the compiled default).
pub const MOONSHOT_CN_BASE_URL_ENV: &str = "KIGI_MOONSHOT_CN_BASE_URL";
/// Base-URL override for moonshot-ai (dev/test escape hatch mirroring
/// `KIGI_CODE_BASE_URL`; production uses the compiled default).
pub const MOONSHOT_AI_BASE_URL_ENV: &str = "KIGI_MOONSHOT_AI_BASE_URL";
/// Env override when set and non-blank, else the compiled default.
fn env_or(var: &str, compiled: &str) -> String {
match std::env::var(var) {
Ok(v) if !v.trim().is_empty() => v,
_ => compiled.to_string(),
}
}
/// Inference dialect a platform speaks. Leaf-safe mirror of the sampler's
/// `ApiBackend` (kigi-models must stay dependency-light); the shell maps it.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum PlatformWireApi {
ChatCompletions,
Responses,
Messages,
}
/// Shape + headers of a platform's model-listing endpoint.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum ListingDialect {
/// `GET {base}/models`, `Authorization: Bearer`, `{data:[{id,...}]}`
/// (the F4 wire contract; kimi extends it with think_efforts etc.).
OpenAi,
/// `GET {base}/models?limit=1000`, `x-api-key` + `anthropic-version`
/// headers, Anthropic's response shape (parsed by
/// [`parse_anthropic_listing`]).
Anthropic,
}
/// ChatCompletions body-adaptation dialect (leaf-safe mirror of the
/// sampler's `ChatCompat`; the shell maps it).
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum PlatformChatCompat {
Kimi,
DeepSeek,
Passthrough,
/// Strict OpenAI-compatible validator (Cerebras, NVIDIA) — strips
/// `stream_options` and private fields.
StrictOpenAi,
/// Mistral: StrictOpenAi plus its exactly-9-alphanumeric tool-call id
/// contract (foreign/OpenAI-style ids are deterministically remapped).
Mistral,
}
/// How a platform's API key rides requests (listing, validation, inference).
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum PlatformKeyHeader {
/// `Authorization: Bearer <key>`.
Bearer,
/// `x-api-key: <key>` plus `anthropic-version` (Anthropic wire).
XApiKey,
}
/// Where a platform's base URL is resolved from.
enum BaseUrlSource {
/// The Kimi Code subscription base, owned by `kigi_env`
/// (honors `KIGI_CODE_BASE_URL`).
KigiEnvCoding,
/// Fixed compiled default with a dev/test env-var override.
EnvOr {
env: &'static str,
default: &'static str,
},
}
/// The interactive login mechanism a `uses_oauth` [`OAuthConfig`] provider
/// drives. `DeviceCode` is the RFC-8628 device flow (xai-grok); `PkceLocalhost`
/// is the authorization-code + PKCE (S256) flow with a loopback callback
/// (Claude Pro/Max). Kimi Code carries neither (its bespoke flow lives in
/// kigi-shell with `oauth: None`).
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum OAuthFlow {
/// RFC-8628 device-code: POST `device_path`, poll `token_path`.
DeviceCode,
/// Authorization-code + PKCE (S256): browser hits `auth_host`+`device_path`
/// (the authorize endpoint); the code returns to a `127.0.0.1:redirect_port`
/// loopback listener answering `redirect_path` (claude `/callback`, codex
/// `/auth/callback`), then is exchanged at `token_host`+`token_path`.
PkceLocalhost {
redirect_port: u16,
redirect_path: &'static str,
},
/// GitHub Copilot two-stage flow (github-copilot): an RFC-8628 device flow
/// on `auth_host` (github.com) mints the DURABLE GitHub token, which is then
/// exchanged at `copilot_exchange` for the SHORT-LIVED copilot session
/// token. The github token is persisted as `refresh_token`; the copilot
/// token as `key`. GitHub's poll returns errors in a `200` body (not `4xx`),
/// so it drives a Copilot-specific poll, not the generic device wire.
GithubDeviceCopilot,
}
/// Body encoding a provider's token endpoint expects for the code-exchange and
/// refresh POSTs. xAI's `/oauth2/token` is form-encoded; Claude's
/// `/v1/oauth/token` is JSON.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum OAuthTokenBody {
/// `application/x-www-form-urlencoded` (xai-grok device wire).
Form,
/// `application/json` (Claude PKCE wire).
Json,
/// GitHub Copilot copilot-token re-mint (github-copilot): "refresh" is NOT
/// a `refresh_token` grant — it is a `GET copilot_internal/v2/token` bearing
/// the durable GitHub token (`refresh_token` field) + editor headers, which
/// re-mints the short-lived copilot token. Dispatched to the Copilot wire.
GithubCopilotExchange,
}
/// Generic OAuth configuration carried by a `uses_oauth` platform whose login
/// is the GENERIC device-code path (xai-grok) or the PKCE-localhost path
/// (claude-pro-max).
///
/// Kimi Code keeps its bespoke device flow (client id, `/api/oauth/*` paths,
/// X-Msh device headers, `kigi_env::oauth_host()`); its `oauth` field stays
/// `None`. All fields here are non-secret wire constants — the access/refresh
/// tokens they mint are NEVER stored in this struct and never logged.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub struct OAuthConfig {
/// OAuth client id sent on every authorize / token call.
pub client_id: &'static str,
/// Authorization-server origin (no trailing slash), e.g. `https://auth.x.ai`
/// (device) or `https://claude.ai` (PKCE authorize host).
pub auth_host: &'static str,
/// Start-endpoint path (POST for `DeviceCode` device-authorization; the
/// browser authorize path for `PkceLocalhost`), relative to `auth_host`.
pub device_path: &'static str,
/// Token-endpoint origin (no trailing slash). Equals `auth_host` for the
/// device wire; for Claude the token host (`https://platform.claude.com`)
/// differs from the authorize host (`https://claude.ai`).
pub token_host: &'static str,
/// Token path (POST) — used for BOTH the initial grant and refresh,
/// relative to `token_host`.
pub token_path: &'static str,
/// OAuth scope string requested at authorization.
pub scope: &'static str,
/// auth.json map key + keyring entry name for this provider's persisted
/// session (e.g. `oauth/xai`).
pub scope_key: &'static str,
/// A non-standard extra form field sent ONLY on the device-authorization
/// request (e.g. `("referrer", "kigi")`). `None` = no extra field.
pub extra_device_field: Option<(&'static str, &'static str)>,
/// Extra query params appended ONLY to the `PkceLocalhost` browser authorize
/// URL (openai-codex's `id_token_add_organizations`, `codex_cli_simplified_
/// flow`, `originator`). Empty for every other config, so their authorize
/// URLs stay byte-identical.
pub authorize_extra: &'static [(&'static str, &'static str)],
/// Interactive login mechanism (device-code vs PKCE-localhost).
pub flow: OAuthFlow,
/// Body encoding the token endpoint expects (form vs JSON).
pub token_body: OAuthTokenBody,
/// Second-stage token-exchange endpoint `(host, path)` for the GitHub
/// Copilot two-stage flow (github-copilot): `("https://api.github.com",
/// "/copilot_internal/v2/token")`. The durable GitHub token is exchanged
/// here for the short-lived copilot token, at BOTH login and every re-mint
/// "refresh". `None` for the standard flows (xai-grok, claude-pro-max),
/// whose refresh is a plain `refresh_token` grant against `token_host`.
pub copilot_exchange: Option<(&'static str, &'static str)>,
/// The access token MUST carry a `chatgpt_account_id` claim (openai-codex):
/// it becomes the `chatgpt-account-id` inference header, so a token without
/// it cannot authorize a request. Login AND every refresh fail fast when the
/// claim is absent. `false` everywhere else — this is an explicit provider
/// fact, NEVER inferred from the token-body encoding (a plain form-encoded
/// token endpoint is the OAuth norm and must not inherit this requirement).
pub requires_chatgpt_account_id: bool,
}
/// xAI / Grok subscription device-code OAuth (ported from Pi
/// `earendil-works/pi` `auth/oauth/xai.ts`).
pub const XAI_OAUTH_CONFIG: OAuthConfig = OAuthConfig {
client_id: "b1a00492-073a-47ea-816f-4c329264a828",
auth_host: "https://auth.x.ai",
device_path: "/oauth2/device/code",
token_host: "https://auth.x.ai",
token_path: "/oauth2/token",
scope: "openid profile email offline_access grok-cli:access api:access",
scope_key: "oauth/xai",
extra_device_field: Some(("referrer", "kigi")),
authorize_extra: &[],
flow: OAuthFlow::DeviceCode,
token_body: OAuthTokenBody::Form,
copilot_exchange: None,
requires_chatgpt_account_id: false,
};
/// Base-URL override for the Claude Pro/Max OAuth channel (dev/test escape
/// hatch). Production defaults to `https://api.anthropic.com/v1`.
pub const CLAUDE_OAUTH_BASE_URL_ENV: &str = "KIGI_CLAUDE_OAUTH_BASE_URL";
/// Claude Pro/Max subscription OAuth (authorization-code + PKCE S256, loopback
/// callback). Authoritative constants from Pi `earendil-works/pi`
/// `auth/oauth/anthropic.ts`: authorize host `https://claude.ai`, token host
/// `https://platform.claude.com` (JSON body), public Claude Code client id.
pub const CLAUDE_OAUTH_CONFIG: OAuthConfig = OAuthConfig {
client_id: "9d1c250a-e61b-44d9-88ed-5944d1962f5e",
auth_host: "https://claude.ai",
device_path: "/oauth/authorize",
token_host: "https://platform.claude.com",
token_path: "/v1/oauth/token",
scope: "org:create_api_key user:profile user:inference user:sessions:claude_code user:mcp_servers user:file_upload",
scope_key: "oauth/claude-pro-max",
extra_device_field: None,
authorize_extra: &[],
flow: OAuthFlow::PkceLocalhost {
redirect_port: 53692,
redirect_path: "/callback",
},
token_body: OAuthTokenBody::Json,
copilot_exchange: None,
requires_chatgpt_account_id: false,
};
/// Base-URL override for the GitHub Copilot inference/listing channel
/// (dev/test escape hatch). Production defaults to the individual-subscription
/// endpoint `https://api.individual.githubcopilot.com`.
pub const COPILOT_BASE_URL_ENV: &str = "KIGI_COPILOT_BASE_URL";
/// GitHub Copilot subscription OAuth (two-stage device flow → copilot-token
/// exchange). Authoritative constants from Pi `earendil-works/pi`
/// `auth/oauth/github-copilot.ts`: the VS Code Copilot Chat public client id,
/// the github.com device endpoints, and the `api.github.com/copilot_internal/
/// v2/token` copilot-token exchange (Stage 2). The device flow mints the
/// durable GitHub token; the exchange re-mints the short-lived copilot token.
pub const COPILOT_OAUTH_CONFIG: OAuthConfig = OAuthConfig {
client_id: "Iv1.b507a08c87ecfe98",
auth_host: "https://github.com",
device_path: "/login/device/code",
token_host: "https://github.com",
token_path: "/login/oauth/access_token",
scope: "read:user",
scope_key: "oauth/github-copilot",
extra_device_field: None,
authorize_extra: &[],
flow: OAuthFlow::GithubDeviceCopilot,
token_body: OAuthTokenBody::GithubCopilotExchange,
copilot_exchange: Some(("https://api.github.com", "/copilot_internal/v2/token")),
requires_chatgpt_account_id: false,
};
/// Base-URL override for the ChatGPT/Codex OAuth inference channel (dev/test
/// escape hatch). Production defaults to the ChatGPT Codex backend
/// `https://chatgpt.com/backend-api/codex`; Kigi's Responses path posts to
/// `{base}/responses`.
pub const CODEX_BASE_URL_ENV: &str = "KIGI_CODEX_BASE_URL";
/// ChatGPT/Codex subscription OAuth (authorization-code + PKCE S256, loopback
/// callback on port 1455 path `/auth/callback`, FORM token body). Authoritative
/// constants from the official Codex CLI + Pi `earendil-works/pi`
/// `auth/oauth/openai-codex.ts`: authorize + token host `https://auth.openai.com`,
/// the `codex_cli_simplified_flow` login client id, and the three authorize-only
/// extra params. Unlike claude the `state` is fresh-random (NOT the verifier),
/// and refresh is a plain `refresh_token` FORM grant (the generic device
/// refresher's `Form` path). The minted `access_token` is a JWT carrying the
/// `chatgpt_account_id` claim consumed at inference time.
pub const CODEX_OAUTH_CONFIG: OAuthConfig = OAuthConfig {
client_id: "app_EMoamEEZ73f0CkXaXp7hrann",
auth_host: "https://auth.openai.com",
device_path: "/oauth/authorize",
token_host: "https://auth.openai.com",
token_path: "/oauth/token",
scope: "openid profile email offline_access",
scope_key: "oauth/openai-codex",
extra_device_field: None,
authorize_extra: &[
("id_token_add_organizations", "true"),
("codex_cli_simplified_flow", "true"),
("originator", "codex_cli_rs"),
],
flow: OAuthFlow::PkceLocalhost {
redirect_port: 1455,
redirect_path: "/auth/callback",
},
token_body: OAuthTokenBody::Form,
copilot_exchange: None,
requires_chatgpt_account_id: true,
};
/// The generic device-code OAuth config for a platform, or `None` for API-key
/// platforms and for Kimi Code (whose bespoke flow uses no generic config).
pub fn oauth_config_for_scope_key(scope_key: &str) -> Option<&'static OAuthConfig> {
PlatformId::ALL
.into_iter()
.find_map(|p| p.oauth().filter(|c| c.scope_key == scope_key))
}
/// One row of the platform registry. All per-platform data lives in these
/// rows; [`PlatformId`] methods only read fields. Adding a platform touches
/// exactly four sites in this file: the enum variant, the `ALL` entry, the
/// `spec()` arm, and the row (plus quirk code where a provider deviates).
/// The `spec()` arm is compiler-enforced; the `ALL` entry is enforced by the
/// `all_covers_every_variant` test — a variant missing from `ALL` would
/// otherwise be silently unparseable and excluded from model sync.
struct PlatformSpec {
/// Wire id (auth method id, managed-model-key prefix, config key, and —
/// for API-key platforms — the auth.json scope the key is stored under).
id: &'static str,
display_name: &'static str,
base_url: BaseUrlSource,
/// True for OAuth-bearer subscription channels.
uses_oauth: bool,
/// Generic device-code OAuth config, for a `uses_oauth` platform whose
/// login is the RFC-8628 device-code path (xai-grok). `None` for API-key
/// platforms AND for Kimi Code (whose bespoke device flow lives in
/// kigi-shell, not this generic config).
oauth: Option<&'static OAuthConfig>,
/// Model-id prefixes admitted from this platform's `/models` listing.
/// `None` = no filtering (listing served pre-filtered).
allowed_model_prefixes: Option<&'static [&'static str]>,
/// Env var names holding this platform's API key, in precedence order
/// (first set, non-blank value wins). Empty for OAuth channels.
///
/// SECURITY: the *values* behind these names must never be logged.
api_key_envs: &'static [&'static str],
/// Short vendor word for login copy ("Paste your {vendor} API key").
vendor: &'static str,
/// Where the user gets an API key (login copy + key-validation errors).
/// `None` for OAuth channels.
console_host: Option<&'static str>,
/// Interactive login-picker label. `None` = fall back to `display_name`.
login_label: Option<&'static str>,
/// This platform's provider id on models.dev, for metadata enrichment.
/// `None` = not covered there (enrichment silently skips).
models_dev_id: Option<&'static str>,
/// True when the platform's `/models` listing itself serves context
/// window / thinking metadata — enrichment (and its network refresh) is
/// skipped entirely for such platforms.
wire_serves_metadata: bool,
/// Inference dialect (mapped to the sampler backend by the shell).
wire_api: PlatformWireApi,
/// Model-listing endpoint shape + headers.
listing: ListingDialect,
/// ChatCompletions body-adaptation dialect (ignored for other backends).
chat_compat: PlatformChatCompat,
/// Key header style for listing/validation/inference.
key_header: PlatformKeyHeader,
/// Restrict the live listing to models the enrichment catalog knows —
/// for providers whose `/models` is polluted with non-chat entries
/// (tts/embeddings/image). Availability still requires the LIVE listing;
/// this only drops listing noise, never adds models.
restrict_to_enriched: bool,
/// Path (relative to base) to hit for API-key VALIDATION at login, when
/// the listing endpoint can't validate. OpenRouter's `/models` is public
/// (200 for any key), so a bad key would false-accept at login; its
/// `/key` endpoint 401s properly. `None` = validate against `/models`.
key_validation_path: Option<&'static str>,
/// A prefix to strip from each live-listing model id before filtering,
/// enrichment lookup, and managed-key formation. Google's OpenAI-compat
/// `/models` returns `models/`-prefixed ids while its chat endpoint (and
/// the models.dev snapshot) use the bare id — stripping canonicalizes to
/// the bare form. `None` = no stripping (the id is used verbatim). The
/// strip is a no-op when the prefix is absent, so it is safe even if a
/// listing returns some ids already bare.
strip_listing_id_prefix: Option<&'static str>,
}
const KIMI_CODE_SPEC: PlatformSpec = PlatformSpec {
id: "kimi-code",
display_name: "Kimi Code",
base_url: BaseUrlSource::KigiEnvCoding,
uses_oauth: true,
// Kimi Code keeps its bespoke device flow (kigi-shell), not the generic
// OAuthConfig path — see architecture decision #1.
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &[],
vendor: "Kimi",
console_host: None,
login_label: None,
models_dev_id: Some("kimi-for-coding"),
wire_serves_metadata: true,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Kimi,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
const MOONSHOT_CN_SPEC: PlatformSpec = PlatformSpec {
id: "moonshot-cn",
display_name: "Moonshot AI Open Platform (moonshot.cn)",
base_url: BaseUrlSource::EnvOr {
env: MOONSHOT_CN_BASE_URL_ENV,
default: "https://api.moonshot.cn/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: Some(&["kimi-k"]),
api_key_envs: &[MOONSHOT_CN_API_KEY_ENV, MOONSHOT_API_KEY_ENV],
vendor: "Moonshot",
console_host: Some("platform.moonshot.cn"),
login_label: Some("Moonshot Open Platform (API key \u{b7} moonshot.cn)"),
models_dev_id: Some("moonshotai-cn"),
wire_serves_metadata: true,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Kimi,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
const MOONSHOT_AI_SPEC: PlatformSpec = PlatformSpec {
id: "moonshot-ai",
display_name: "Moonshot AI Open Platform (moonshot.ai)",
base_url: BaseUrlSource::EnvOr {
env: MOONSHOT_AI_BASE_URL_ENV,
default: "https://api.moonshot.ai/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: Some(&["kimi-k"]),
api_key_envs: &[MOONSHOT_AI_API_KEY_ENV, MOONSHOT_API_KEY_ENV],
vendor: "Moonshot",
console_host: Some("platform.moonshot.ai"),
login_label: Some("Moonshot Open Platform (API key \u{b7} moonshot.ai)"),
models_dev_id: Some("moonshotai"),
wire_serves_metadata: true,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Kimi,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for OpenAI (dev/test escape hatch).
pub const OPENAI_BASE_URL_ENV: &str = "KIGI_OPENAI_BASE_URL";
const OPENAI_SPEC: PlatformSpec = PlatformSpec {
id: "openai",
display_name: "OpenAI",
base_url: BaseUrlSource::EnvOr {
env: OPENAI_BASE_URL_ENV,
default: "https://api.openai.com/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["OPENAI_API_KEY"],
vendor: "OpenAI",
console_host: Some("platform.openai.com"),
login_label: Some("OpenAI (API key)"),
models_dev_id: Some("openai"),
// GET /v1/models returns bare ids only (no context/thinking metadata)
// and is polluted with tts/embeddings/image entries.
wire_serves_metadata: false,
wire_api: PlatformWireApi::Responses,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: true,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for Anthropic (dev/test escape hatch).
pub const ANTHROPIC_BASE_URL_ENV: &str = "KIGI_ANTHROPIC_BASE_URL";
const ANTHROPIC_SPEC: PlatformSpec = PlatformSpec {
id: "anthropic",
display_name: "Anthropic",
base_url: BaseUrlSource::EnvOr {
env: ANTHROPIC_BASE_URL_ENV,
default: "https://api.anthropic.com/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["ANTHROPIC_API_KEY"],
vendor: "Anthropic",
console_host: Some("console.anthropic.com"),
login_label: Some("Anthropic (API key)"),
models_dev_id: Some("anthropic"),
// The 2026 /v1/models serves capabilities + max_input_tokens, but the
// adapter maps only what's present — enrichment fills gaps (wire wins).
wire_serves_metadata: false,
wire_api: PlatformWireApi::Messages,
listing: ListingDialect::Anthropic,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::XApiKey,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for DeepSeek (dev/test escape hatch).
pub const DEEPSEEK_BASE_URL_ENV: &str = "KIGI_DEEPSEEK_BASE_URL";
const DEEPSEEK_SPEC: PlatformSpec = PlatformSpec {
id: "deepseek",
display_name: "DeepSeek",
base_url: BaseUrlSource::EnvOr {
env: DEEPSEEK_BASE_URL_ENV,
// No /v1: chat rides {base}/chat/completions, listing {base}/models
// (official docs).
default: "https://api.deepseek.com",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["DEEPSEEK_API_KEY"],
vendor: "DeepSeek",
console_host: Some("platform.deepseek.com"),
login_label: Some("DeepSeek (API key)"),
models_dev_id: Some("deepseek"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::DeepSeek,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for Groq (dev/test escape hatch).
pub const GROQ_BASE_URL_ENV: &str = "KIGI_GROQ_BASE_URL";
const GROQ_SPEC: PlatformSpec = PlatformSpec {
id: "groq",
display_name: "Groq",
base_url: BaseUrlSource::EnvOr {
env: GROQ_BASE_URL_ENV,
default: "https://api.groq.com/openai/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["GROQ_API_KEY"],
vendor: "Groq",
console_host: Some("console.groq.com"),
login_label: Some("Groq (API key)"),
models_dev_id: Some("groq"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// The listing carries whisper/tts entries; keep tool-calling chat
// models only.
restrict_to_enriched: true,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for Mistral (dev/test escape hatch).
pub const MISTRAL_BASE_URL_ENV: &str = "KIGI_MISTRAL_BASE_URL";
const MISTRAL_SPEC: PlatformSpec = PlatformSpec {
id: "mistral",
display_name: "Mistral",
base_url: BaseUrlSource::EnvOr {
env: MISTRAL_BASE_URL_ENV,
default: "https://api.mistral.ai/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["MISTRAL_API_KEY"],
vendor: "Mistral",
console_host: Some("console.mistral.ai"),
login_label: Some("Mistral (API key)"),
models_dev_id: Some("mistral"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
// Mistral's strict validator 422s on `stream_options`, its reasoning
// models return array content, and tool-call ids must be EXACTLY nine
// `[a-zA-Z0-9]` chars — the Mistral dialect strips stream_options and
// deterministically remaps non-conforming (foreign/OpenAI-style) ids;
// the response deserializer handles arrays universally.
chat_compat: PlatformChatCompat::Mistral,
key_header: PlatformKeyHeader::Bearer,
// The listing carries embed/moderation/OCR entries; keep tool-calling
// chat models only.
restrict_to_enriched: true,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for Fireworks (dev/test escape hatch).
pub const FIREWORKS_BASE_URL_ENV: &str = "KIGI_FIREWORKS_BASE_URL";
const FIREWORKS_SPEC: PlatformSpec = PlatformSpec {
id: "fireworks",
display_name: "Fireworks AI",
base_url: BaseUrlSource::EnvOr {
env: FIREWORKS_BASE_URL_ENV,
// Inference plane (note the /inference/v1 path, not /v1).
default: "https://api.fireworks.ai/inference/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["FIREWORKS_API_KEY"],
vendor: "Fireworks",
console_host: Some("fireworks.ai"),
login_label: Some("Fireworks AI (API key)"),
models_dev_id: Some("fireworks-ai"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// The inference /models listing can include embedding/non-chat models;
// keep tool-calling enrichment-known models only.
restrict_to_enriched: true,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// Base-URL override for Google Gemini (dev/test escape hatch).
pub const GOOGLE_BASE_URL_ENV: &str = "KIGI_GOOGLE_BASE_URL";
const GOOGLE_SPEC: PlatformSpec = PlatformSpec {
id: "google",
display_name: "Google Gemini",
base_url: BaseUrlSource::EnvOr {
env: GOOGLE_BASE_URL_ENV,
// Gemini's OpenAI-compatibility shim (bare model ids, Bearer key).
default: "https://generativelanguage.googleapis.com/v1beta/openai",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["GEMINI_API_KEY"],
vendor: "Google",
console_host: Some("aistudio.google.com"),
login_label: Some("Google Gemini (API key)"),
models_dev_id: Some("google"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// The compat listing carries embedding/tts/image models; keep
// tool-calling enrichment-known chat models only.
restrict_to_enriched: true,
key_validation_path: None,
strip_listing_id_prefix: Some("models/"),
};
/// Base-URL override for OpenRouter (dev/test escape hatch).
pub const OPENROUTER_BASE_URL_ENV: &str = "KIGI_OPENROUTER_BASE_URL";
const OPENROUTER_SPEC: PlatformSpec = PlatformSpec {
id: "openrouter",
display_name: "OpenRouter",
base_url: BaseUrlSource::EnvOr {
env: OPENROUTER_BASE_URL_ENV,
// Note /api/v1, not /v1.
default: "https://openrouter.ai/api/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["OPENROUTER_API_KEY"],
vendor: "OpenRouter",
console_host: Some("openrouter.ai"),
login_label: Some("OpenRouter (API key)"),
// OpenRouter's public /models serves context_length for every model, so
// enrichment is neither needed nor fetched (verified live: 340/340 carry
// a top-level context_length). Not a models.dev provider here.
models_dev_id: None,
wire_serves_metadata: true,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: Some("/key"),
strip_listing_id_prefix: None,
};
/// Base-URL override for Together AI (dev/test escape hatch).
pub const TOGETHER_BASE_URL_ENV: &str = "KIGI_TOGETHER_BASE_URL";
const TOGETHER_SPEC: PlatformSpec = PlatformSpec {
id: "together",
display_name: "Together AI",
base_url: BaseUrlSource::EnvOr {
env: TOGETHER_BASE_URL_ENV,
default: "https://api.together.xyz/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["TOGETHER_API_KEY"],
vendor: "Together",
console_host: Some("api.together.xyz"),
login_label: Some("Together AI (API key)"),
models_dev_id: Some("togetherai"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
// Together's /v1/models is a BARE JSON array (parse_openai_listing is
// tolerant), and it mixes chat/embedding/rerank/image types; keep
// tool-calling enrichment-known chat models only.
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
/// Base-URL override for Cerebras (dev/test escape hatch).
pub const CEREBRAS_BASE_URL_ENV: &str = "KIGI_CEREBRAS_BASE_URL";
const CEREBRAS_SPEC: PlatformSpec = PlatformSpec {
id: "cerebras",
display_name: "Cerebras",
base_url: BaseUrlSource::EnvOr {
env: CEREBRAS_BASE_URL_ENV,
default: "https://api.cerebras.ai/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["CEREBRAS_API_KEY"],
vendor: "Cerebras",
console_host: Some("cloud.cerebras.ai"),
login_label: Some("Cerebras (API key)"),
models_dev_id: Some("cerebras"),
// /models is minimal (id only, no context) → enrichment supplies context
// + effort menus. The catalog is all chat LLMs (no embedding/tts
// pollution), so keep every live model and enrich the known ones.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
// Cerebras uses strict additionalProperties:false validation (400s on
// out-of-schema fields like store/thinking); strip stream_options.
chat_compat: PlatformChatCompat::StrictOpenAi,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: false,
};
/// Base-URL override for NVIDIA NIM (dev/test escape hatch).
pub const NVIDIA_BASE_URL_ENV: &str = "KIGI_NVIDIA_BASE_URL";
const NVIDIA_SPEC: PlatformSpec = PlatformSpec {
id: "nvidia",
display_name: "NVIDIA NIM",
base_url: BaseUrlSource::EnvOr {
env: NVIDIA_BASE_URL_ENV,
default: "https://integrate.api.nvidia.com/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["NVIDIA_API_KEY"],
vendor: "NVIDIA",
console_host: Some("build.nvidia.com"),
login_label: Some("NVIDIA NIM (API key)"),
models_dev_id: Some("nvidia"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
// NIM exposes raw vLLM behavior; stream_options support varies per model
// and some 4xx on it, so strip it (StrictOpenAi) to keep streaming
// working across the fleet.
chat_compat: PlatformChatCompat::StrictOpenAi,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
// The listing mixes chat/embedding/rerank/vision/image models; keep
// tool-calling enrichment-known chat models only.
restrict_to_enriched: true,
};
/// Base-URL override for Vercel AI Gateway (dev/test escape hatch).
pub const VERCEL_BASE_URL_ENV: &str = "KIGI_VERCEL_BASE_URL";
const VERCEL_SPEC: PlatformSpec = PlatformSpec {
id: "vercel-ai-gateway",
display_name: "Vercel AI Gateway",
base_url: BaseUrlSource::EnvOr {
env: VERCEL_BASE_URL_ENV,
default: "https://ai-gateway.vercel.sh/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["AI_GATEWAY_API_KEY"],
vendor: "Vercel",
console_host: Some("vercel.com"),
login_label: Some("Vercel AI Gateway (API key)"),
models_dev_id: Some("vercel"),
// Vercel's /models serves rich metadata but under `context_window` (not
// the WireModel `context_length`), so take context from models.dev
// enrichment instead; restrict to tool-calling chat models (the gateway
// lists embedding/image/rerank types too). Ids are creator/model,
// byte-matching the models.dev "vercel" keys.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// /models is PUBLIC (200 for any key), so validate against /credits,
// which 401s for a bad key.
key_validation_path: Some("/credits"),
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const XAI_BASE_URL_ENV: &str = "KIGI_XAI_BASE_URL";
const XAI_SPEC: PlatformSpec = PlatformSpec {
id: "xai",
display_name: "xAI (Grok)",
base_url: BaseUrlSource::EnvOr {
env: XAI_BASE_URL_ENV,
default: "https://api.x.ai/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
// NOTE: `XAI_API_KEY` is ALSO read as a legacy fallback by the house BYOK
// path (`read_xai_api_key_env`), whose primary env is `KIGI_API_KEY`.
// Here it is the x.ai/Grok provider key (its canonical ecosystem name).
api_key_envs: &["XAI_API_KEY"],
vendor: "xAI",
console_host: Some("console.x.ai"),
login_label: Some("xAI (Grok) (API key)"),
models_dev_id: Some("xai"),
// /v1/models is minimal (ids only; rich metadata lives on the non-standard
// /v1/language-models), so take context/limits from models.dev enrichment.
// The /v1/models ids match the models.dev "xai" keys byte-for-byte
// (grok-4.5, grok-4.20-0309-reasoning, ...); restrict to tool-calling chat
// models to drop the grok-imagine-* image/video generators.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// /v1/models requires auth (401 without a key), so it doubles as the key
// validator; no separate public endpoint needed.
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
/// Base-URL override for the xAI Grok subscription (OAuth) platform. Its own
/// env (distinct from `KIGI_XAI_BASE_URL`) so a mock can target it
/// independently of the API-key `xai` row.
pub const XAI_GROK_BASE_URL_ENV: &str = "KIGI_XAI_GROK_BASE_URL";
const XAI_GROK_SPEC: PlatformSpec = PlatformSpec {
id: "xai-grok",
display_name: "xAI (Grok subscription)",
// Same wire as the API-key `xai` row (api.x.ai/v1), but reached with an
// OAuth subscription bearer instead of an API key.
base_url: BaseUrlSource::EnvOr {
env: XAI_GROK_BASE_URL_ENV,
default: "https://api.x.ai/v1",
},
uses_oauth: true,
oauth: Some(&XAI_OAUTH_CONFIG),
allowed_model_prefixes: None,
// OAuth channel: no API key envs (the device-flow session is the bearer).
api_key_envs: &[],
vendor: "xAI",
console_host: Some("x.ai"),
login_label: Some("xAI Grok (subscription)"),
models_dev_id: Some("xai"),
// Same as XAI_SPEC: bare-id /v1/models, enrich from models.dev "xai",
// restrict to tool-calling chat models (drops the grok-imagine-* media
// generators).
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const QWEN_TOKEN_PLAN_BASE_URL_ENV: &str = "KIGI_QWEN_TOKEN_PLAN_BASE_URL";
const QWEN_TOKEN_PLAN_SPEC: PlatformSpec = PlatformSpec {
id: "qwen-token-plan",
display_name: "Qwen Token Plan",
base_url: BaseUrlSource::EnvOr {
env: QWEN_TOKEN_PLAN_BASE_URL_ENV,
default: "https://token-plan.ap-southeast-1.maas.aliyuncs.com/compatible-mode/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["QWEN_TOKEN_PLAN_API_KEY"],
vendor: "Alibaba",
console_host: Some("modelstudio.console.alibabacloud.com"),
login_label: Some("Qwen Token Plan (API key)"),
models_dev_id: Some("alibaba-token-plan"),
// DashScope compatible-mode /models is auth-gated (no wire metadata), so
// take context/limits from the models.dev "alibaba-token-plan" snapshot;
// restrict to tool-calling chat models to drop the qwen-image / wan image
// generators the token plan also lists.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// /models requires auth (401 without a key), so it doubles as the key
// validator.
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const QWEN_TOKEN_PLAN_CN_BASE_URL_ENV: &str = "KIGI_QWEN_TOKEN_PLAN_CN_BASE_URL";
const QWEN_TOKEN_PLAN_CN_SPEC: PlatformSpec = PlatformSpec {
id: "qwen-token-plan-cn",
display_name: "Qwen Token Plan (China)",
base_url: BaseUrlSource::EnvOr {
env: QWEN_TOKEN_PLAN_CN_BASE_URL_ENV,
default: "https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["QWEN_TOKEN_PLAN_CN_API_KEY"],
vendor: "Alibaba",
console_host: Some("bailian.console.aliyun.com"),
login_label: Some("Qwen Token Plan China (API key)"),
models_dev_id: Some("alibaba-token-plan-cn"),
// China endpoint; same DashScope compatible-mode shape as the global plan.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
const KIMI_CODING_SPEC: PlatformSpec = PlatformSpec {
id: "kimi-coding",
display_name: "Kimi For Coding",
// Same endpoint as the OAuth `kimi-code` platform (KIGI_CODE_BASE_URL
// override), but authenticated with a static KIMI_API_KEY instead of the
// device flow. Kimi's /coding/v1/models serves its own metadata and the
// Kimi thinking dialect applies verbatim.
base_url: BaseUrlSource::KigiEnvCoding,
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["KIMI_API_KEY"],
vendor: "Kimi",
console_host: Some("www.kimi.com"),
login_label: Some("Kimi For Coding (API key)"),
models_dev_id: Some("kimi-for-coding"),
wire_serves_metadata: true,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Kimi,
key_header: PlatformKeyHeader::Bearer,
// /coding/v1/models requires auth (401 without a key), so it doubles as
// the key validator.
key_validation_path: None,
restrict_to_enriched: false,
strip_listing_id_prefix: None,
};
pub const ZAI_BASE_URL_ENV: &str = "KIGI_ZAI_BASE_URL";
const ZAI_SPEC: PlatformSpec = PlatformSpec {
id: "zai",
display_name: "Z.AI",
// Z.AI coding plan (GLM). Per Pi (earendil-works/pi) this is plain
// OpenAI-compatible chat completions — no special thinking dialect.
base_url: BaseUrlSource::EnvOr {
env: ZAI_BASE_URL_ENV,
default: "https://api.z.ai/api/coding/paas/v4",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["ZAI_API_KEY"],
vendor: "Z.AI",
console_host: Some("z.ai"),
login_label: Some("Z.AI (API key)"),
models_dev_id: Some("zai-coding-plan"),
// /models auth-gated → validator; enrichment from models.dev; restrict to
// tool-calling GLM chat models. Live ids match the snapshot keys
// (glm-4.7, glm-5-turbo, glm-5.2, ...) byte-for-byte (verified vs Pi).
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const ZAI_CODING_CN_BASE_URL_ENV: &str = "KIGI_ZAI_CODING_CN_BASE_URL";
const ZAI_CODING_CN_SPEC: PlatformSpec = PlatformSpec {
id: "zai-coding-cn",
display_name: "Z.AI Coding (China)",
// Zhipu/BigModel-hosted CN coding plan; same OpenAI-compatible shape.
base_url: BaseUrlSource::EnvOr {
env: ZAI_CODING_CN_BASE_URL_ENV,
default: "https://open.bigmodel.cn/api/coding/paas/v4",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["ZAI_CODING_CN_API_KEY"],
vendor: "Z.AI",
console_host: Some("open.bigmodel.cn"),
login_label: Some("Z.AI Coding China (API key)"),
models_dev_id: Some("zhipuai-coding-plan"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const XIAOMI_BASE_URL_ENV: &str = "KIGI_XIAOMI_BASE_URL";
const XIAOMI_SPEC: PlatformSpec = PlatformSpec {
id: "xiaomi",
display_name: "Xiaomi MiMo",
// Xiaomi MiMo. Per Pi this is plain OpenAI-compatible chat completions.
base_url: BaseUrlSource::EnvOr {
env: XIAOMI_BASE_URL_ENV,
default: "https://api.xiaomimimo.com/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["XIAOMI_API_KEY"],
vendor: "Xiaomi",
console_host: Some("xiaomimimo.com"),
login_label: Some("Xiaomi MiMo (API key)"),
models_dev_id: Some("xiaomi"),
// /models auth-gated → validator; enrichment from models.dev; restrict to
// tool-calling chat models. Live ids match the snapshot keys (mimo-v2.5,
// mimo-v2-pro, ...) byte-for-byte (verified vs Pi).
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const XIAOMI_TOKEN_PLAN_CN_BASE_URL_ENV: &str = "KIGI_XIAOMI_TOKEN_PLAN_CN_BASE_URL";
const XIAOMI_TOKEN_PLAN_CN_SPEC: PlatformSpec = PlatformSpec {
id: "xiaomi-token-plan-cn",
display_name: "Xiaomi Token Plan (China)",
base_url: BaseUrlSource::EnvOr {
env: XIAOMI_TOKEN_PLAN_CN_BASE_URL_ENV,
default: "https://token-plan-cn.xiaomimimo.com/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["XIAOMI_TOKEN_PLAN_CN_API_KEY"],
vendor: "Xiaomi",
console_host: Some("xiaomimimo.com"),
login_label: Some("Xiaomi Token Plan China (API key)"),
models_dev_id: Some("xiaomi-token-plan-cn"),
// The CN token plan listing also carries mimo TTS models (tool_call=false)
// → restrict drops them, keeping only the tool-calling chat models.
wire_serves_metadata: false,
wire_api: PlatformWireApi::ChatCompletions,
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: true,
};
pub const MINIMAX_BASE_URL_ENV: &str = "KIGI_MINIMAX_BASE_URL";
const MINIMAX_SPEC: PlatformSpec = PlatformSpec {
id: "minimax",
display_name: "MiniMax",
// Per Pi, MiniMax is driven through its Anthropic-compatible surface
// (baseUrl .../anthropic). Kigi appends bare paths, so the base carries
// the /v1 suffix: listing → .../anthropic/v1/models?limit=1000, inference
// → .../anthropic/v1/messages. Reuses the Anthropic Messages machinery.
base_url: BaseUrlSource::EnvOr {
env: MINIMAX_BASE_URL_ENV,
default: "https://api.minimax.io/anthropic/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["MINIMAX_API_KEY"],
vendor: "MiniMax",
console_host: Some("platform.minimax.io"),
login_label: Some("MiniMax (API key)"),
models_dev_id: Some("minimax"),
// Anthropic listing (x-api-key + anthropic-version) serves the id list;
// enrichment fills context. Catalog is clean (7 MiniMax-M* chat models),
// so no restriction — and restrict=false keeps launch-day models that
// models.dev has not indexed yet.
wire_serves_metadata: false,
wire_api: PlatformWireApi::Messages,
listing: ListingDialect::Anthropic,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::XApiKey,
// /anthropic/v1/models requires x-api-key (401 without), so it validates.
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: false,
};
pub const MINIMAX_CN_BASE_URL_ENV: &str = "KIGI_MINIMAX_CN_BASE_URL";
const MINIMAX_CN_SPEC: PlatformSpec = PlatformSpec {
id: "minimax-cn",
display_name: "MiniMax (China)",
base_url: BaseUrlSource::EnvOr {
env: MINIMAX_CN_BASE_URL_ENV,
default: "https://api.minimaxi.com/anthropic/v1",
},
uses_oauth: false,
oauth: None,
allowed_model_prefixes: None,
api_key_envs: &["MINIMAX_CN_API_KEY"],
vendor: "MiniMax",
console_host: Some("platform.minimaxi.com"),
login_label: Some("MiniMax China (API key)"),
models_dev_id: Some("minimax-cn"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::Messages,
listing: ListingDialect::Anthropic,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::XApiKey,
key_validation_path: None,
strip_listing_id_prefix: None,
restrict_to_enriched: false,
};
/// Claude Pro/Max subscription via PKCE-localhost OAuth. Reaches
/// `api.anthropic.com` with an OAuth `sk-ant-oat…` bearer (NOT an API key) —
/// same Anthropic Messages + listing wire as the API-key `anthropic` row, plus
/// the OAuth identity headers + "You are Claude Code" system prefix (gated on
/// this platform's OAuth path in the sampler/fetch, so the API-key rows stay
/// byte-identical).
const CLAUDE_PRO_MAX_SPEC: PlatformSpec = PlatformSpec {
id: "claude-pro-max",
display_name: "Claude Pro/Max",
// WITH /v1 so listing → /v1/models and inference → /v1/messages, matching
// ANTHROPIC_SPEC's base handling.
base_url: BaseUrlSource::EnvOr {
env: CLAUDE_OAUTH_BASE_URL_ENV,
default: "https://api.anthropic.com/v1",
},
uses_oauth: true,
oauth: Some(&CLAUDE_OAUTH_CONFIG),
allowed_model_prefixes: None,
// OAuth channel: no API key envs (the PKCE session is the bearer).
api_key_envs: &[],
vendor: "Anthropic",
console_host: None,
login_label: Some("Claude Pro/Max (subscription)"),
models_dev_id: Some("anthropic"),
wire_serves_metadata: false,
wire_api: PlatformWireApi::Messages,
listing: ListingDialect::Anthropic,
// Passthrough is ignored for the Messages backend.
chat_compat: PlatformChatCompat::Passthrough,
// OAuth uses Authorization: Bearer, NOT x-api-key.
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
const GITHUB_COPILOT_SPEC: PlatformSpec = PlatformSpec {
id: "github-copilot",
display_name: "GitHub Copilot",
base_url: BaseUrlSource::EnvOr {
env: COPILOT_BASE_URL_ENV,
default: "https://api.individual.githubcopilot.com",
},
uses_oauth: true,
oauth: Some(&COPILOT_OAUTH_CONFIG),
allowed_model_prefixes: None,
// OAuth channel: no API key envs (the copilot session is the bearer).
api_key_envs: &[],
vendor: "GitHub",
console_host: Some("github.com"),
login_label: Some("GitHub Copilot (subscription)"),
models_dev_id: Some("github-copilot"),
// Copilot /models serves availability flags (model_picker_enabled/policy/
// tool_calls), NOT context/thinking metadata — enrich from models.dev.
wire_serves_metadata: false,
// ChatCompletions ONLY: the catalog is filtered (see
// `parse_github_copilot_listing`) to the openai-completions-served models.
// The claude-4.x/5.x (messages) and gpt-5/oswe/mai- (responses-only) models
// this endpoint also lists are EXCLUDED — Kigi is one-wire-per-platform and
// per-model wire routing is deferred (documented limitation).
wire_api: PlatformWireApi::ChatCompletions,
// OpenAI-shape listing endpoint, but with Copilot-specific availability
// fields — parsed by `parse_github_copilot_listing`, gated on the platform.
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
// The copilot listing filter + the wire-compat filter govern the catalog,
// NOT the enrichment membership (a live but enrichment-lagging model stays).
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
const OPENAI_CODEX_SPEC: PlatformSpec = PlatformSpec {
id: "openai-codex",
display_name: "ChatGPT (Codex)",
// Kigi's Responses path posts to `{base}/responses`; the codex backend
// serves it at `.../codex/responses`, so the base carries the `/codex` tail.
base_url: BaseUrlSource::EnvOr {
env: CODEX_BASE_URL_ENV,
default: "https://chatgpt.com/backend-api/codex",
},
uses_oauth: true,
oauth: Some(&CODEX_OAUTH_CONFIG),
allowed_model_prefixes: None,
// OAuth channel: no API key envs (the PKCE session is the bearer).
api_key_envs: &[],
vendor: "OpenAI",
console_host: Some("chatgpt.com"),
login_label: Some("ChatGPT Plus/Pro (Codex)"),
// HARDCODED catalog (see `hardcoded_catalog`): NOT enriched from models.dev
// and NOT live-fetched — OpenAI exposes no stable public models endpoint for
// this backend.
models_dev_id: None,
// The catalog is compiled-in and already carries context/thinking metadata,
// so enrichment (and its network refresh) is skipped entirely.
wire_serves_metadata: true,
wire_api: PlatformWireApi::Responses,
// Unused: the catalog does NOT come from a live `/models` listing (the
// fetch path short-circuits to `hardcoded_catalog`).
listing: ListingDialect::OpenAi,
chat_compat: PlatformChatCompat::Passthrough,
key_header: PlatformKeyHeader::Bearer,
restrict_to_enriched: false,
key_validation_path: None,
strip_listing_id_prefix: None,
};
/// The platform registry. Platforms are compiled-in spec rows; there is no
/// dynamic provider registration (PRD F2).
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash, PartialOrd, Ord)]
pub enum PlatformId {
/// Kimi Code subscription (OAuth bearer from the F1 device flow).
KimiCode,
/// Moonshot AI open platform, api.moonshot.cn (API key).
MoonshotCn,
/// Moonshot AI open platform, api.moonshot.ai (API key).
MoonshotAi,
/// OpenAI platform API (API key, Responses dialect).
OpenAi,
/// Anthropic platform API (API key, Messages dialect).
Anthropic,
/// DeepSeek platform API (API key, ChatCompletions dialect).
DeepSeek,
/// Groq platform API (API key, OpenAI-compatible ChatCompletions).
Groq,
/// Mistral platform API (API key, OpenAI-compatible ChatCompletions).
Mistral,
/// Fireworks AI platform API (API key, OpenAI-compatible ChatCompletions).
Fireworks,
/// Google Gemini platform API (API key, OpenAI-compatibility shim).
Google,
/// OpenRouter meta-provider (API key, wire-served metadata).
OpenRouter,
/// Together AI platform API (API key, bare-array listing).
Together,
/// Cerebras platform API (API key, OpenAI-compatible ChatCompletions).
Cerebras,
/// NVIDIA NIM platform API (API key, OpenAI-compatible ChatCompletions).
Nvidia,
/// Vercel AI Gateway (API key, wire-listed with models.dev enrichment).
Vercel,
/// xAI Grok platform (API key, OpenAI-compatible ChatCompletions).
Xai,
/// Alibaba Qwen Token Plan, global (API key, DashScope compatible-mode).
QwenTokenPlan,
/// Alibaba Qwen Token Plan, China (API key, DashScope compatible-mode).
QwenTokenPlanCn,
/// Kimi For Coding via a static KIMI_API_KEY (same endpoint as `KimiCode`).
KimiCoding,
/// Z.AI coding plan, global (API key, OpenAI-compatible GLM).
Zai,
/// Z.AI coding plan, China / Zhipu BigModel (API key, OpenAI-compatible).
ZaiCodingCn,
/// Xiaomi MiMo, global (API key, OpenAI-compatible).
Xiaomi,
/// Xiaomi Token Plan, China (API key, OpenAI-compatible).
XiaomiTokenPlanCn,
/// MiniMax, global (API key, Anthropic-compatible Messages).
Minimax,
/// MiniMax, China (API key, Anthropic-compatible Messages).
MinimaxCn,
/// xAI Grok subscription via device-code OAuth (same wire as `Xai`).
XaiGrok,
/// Claude Pro/Max subscription via PKCE-localhost OAuth (Anthropic Messages
/// wire reached with an OAuth bearer instead of an API key).
ClaudeProMax,
/// GitHub Copilot subscription via the two-stage device-code OAuth flow
/// (ChatCompletions wire reached with the short-lived copilot token).
GithubCopilot,
/// ChatGPT Plus/Pro subscription via PKCE-localhost OAuth against the
/// ChatGPT Codex backend (Responses wire reached with an OAuth bearer +
/// the `chatgpt-account-id` JWT claim). HARDCODED catalog, no live listing.
OpenaiCodex,
}
impl PlatformId {
/// All platforms, in catalog precedence order: the subscription channel
/// first so "default model = first list item" favors it when present.
pub const ALL: [PlatformId; 29] = [
Self::KimiCode,
Self::MoonshotCn,
Self::MoonshotAi,
Self::OpenAi,
Self::Anthropic,
Self::DeepSeek,
Self::Groq,
Self::Mistral,
Self::Fireworks,
Self::Google,
Self::OpenRouter,
Self::Together,
Self::Cerebras,
Self::Nvidia,
Self::Vercel,
Self::Xai,
Self::QwenTokenPlan,
Self::QwenTokenPlanCn,
Self::KimiCoding,
Self::Zai,
Self::ZaiCodingCn,
Self::Xiaomi,
Self::XiaomiTokenPlanCn,
Self::Minimax,
Self::MinimaxCn,
Self::XaiGrok,
Self::ClaudeProMax,
Self::GithubCopilot,
Self::OpenaiCodex,
];
/// The registry row backing this platform (single source of per-platform
/// data; every accessor below reads it).
const fn spec(self) -> &'static PlatformSpec {
match self {
Self::KimiCode => &KIMI_CODE_SPEC,
Self::MoonshotCn => &MOONSHOT_CN_SPEC,
Self::MoonshotAi => &MOONSHOT_AI_SPEC,
Self::OpenAi => &OPENAI_SPEC,
Self::Anthropic => &ANTHROPIC_SPEC,
Self::DeepSeek => &DEEPSEEK_SPEC,
Self::Groq => &GROQ_SPEC,
Self::Mistral => &MISTRAL_SPEC,
Self::Fireworks => &FIREWORKS_SPEC,
Self::Google => &GOOGLE_SPEC,
Self::OpenRouter => &OPENROUTER_SPEC,
Self::Together => &TOGETHER_SPEC,
Self::Cerebras => &CEREBRAS_SPEC,
Self::Nvidia => &NVIDIA_SPEC,
Self::Vercel => &VERCEL_SPEC,
Self::Xai => &XAI_SPEC,
Self::QwenTokenPlan => &QWEN_TOKEN_PLAN_SPEC,
Self::QwenTokenPlanCn => &QWEN_TOKEN_PLAN_CN_SPEC,
Self::KimiCoding => &KIMI_CODING_SPEC,
Self::Zai => &ZAI_SPEC,
Self::ZaiCodingCn => &ZAI_CODING_CN_SPEC,
Self::Xiaomi => &XIAOMI_SPEC,
Self::XiaomiTokenPlanCn => &XIAOMI_TOKEN_PLAN_CN_SPEC,
Self::Minimax => &MINIMAX_SPEC,
Self::MinimaxCn => &MINIMAX_CN_SPEC,
Self::XaiGrok => &XAI_GROK_SPEC,
Self::ClaudeProMax => &CLAUDE_PRO_MAX_SPEC,
Self::GithubCopilot => &GITHUB_COPILOT_SPEC,
Self::OpenaiCodex => &OPENAI_CODEX_SPEC,
}
}
pub fn as_str(self) -> &'static str {
self.spec().id
}
pub fn parse(s: &str) -> Option<Self> {
Self::ALL.into_iter().find(|p| p.spec().id == s)
}
pub fn display_name(self) -> &'static str {
self.spec().display_name
}
/// Inference/model-listing base URL. The subscription base honors the
/// `KIGI_CODE_BASE_URL` override via [`kigi_env::coding_api_base_url`];
/// API-key platform bases are fixed in production, with a per-platform
/// env var (e.g. `KIGI_MOONSHOT_{CN,AI}_BASE_URL`) as dev/test override.
pub fn base_url(self) -> String {
match self.spec().base_url {
BaseUrlSource::KigiEnvCoding => kigi_env::coding_api_base_url(),
BaseUrlSource::EnvOr { env, default } => env_or(env, default),
}
}
/// True for OAuth-bearer subscription channels.
pub fn uses_oauth(self) -> bool {
self.spec().uses_oauth
}
/// Generic device-code OAuth config for this platform, or `None` for
/// API-key platforms and for Kimi Code (bespoke flow). A `Some` here means
/// `uses_oauth()` is also true; the fetch/base-url routing treats these
/// providers as their own `base_url()` (not the Kimi `proxy_url()`).
pub fn oauth(self) -> Option<&'static OAuthConfig> {
self.spec().oauth
}
/// Model-id prefixes admitted from this platform's `/models` listing.
/// `None` = no filtering (subscription listing is served pre-filtered).
pub fn allowed_model_prefixes(self) -> Option<&'static [&'static str]> {
self.spec().allowed_model_prefixes
}
/// Env var names holding this platform's API key, in precedence order
/// (first set, non-blank value wins). Empty for the OAuth channel.
///
/// SECURITY: the *values* behind these names must never be logged.
pub fn api_key_env_names(self) -> &'static [&'static str] {
self.spec().api_key_envs
}
/// Managed catalog key for a model served by this platform:
/// `{platform_id}/{model_id}` (kimi-cli `managed_model_key`).
pub fn managed_model_key(self, model_id: &str) -> String {
format!("{}/{model_id}", self.as_str())
}
/// Short vendor word for login copy ("Paste your {vendor} API key").
pub fn vendor(self) -> &'static str {
self.spec().vendor
}
/// Console host where the user obtains an API key, for login copy and
/// key-validation errors. `None` for OAuth channels.
pub fn console_host(self) -> Option<&'static str> {
self.spec().console_host
}
/// Label for the interactive login picker (falls back to the display
/// name when the row doesn't override it).
pub fn login_label(self) -> &'static str {
self.spec().login_label.unwrap_or(self.spec().display_name)
}
/// This platform's provider id on models.dev (metadata enrichment).
pub fn models_dev_id(self) -> Option<&'static str> {
self.spec().models_dev_id
}
/// True when the live `/models` wire serves metadata itself — enrichment
/// and its network refresh are skipped for such platforms.
pub fn wire_serves_metadata(self) -> bool {
self.spec().wire_serves_metadata
}
/// Inference dialect this platform speaks (shell maps to `ApiBackend`).
pub fn wire_api(self) -> PlatformWireApi {
self.spec().wire_api
}
/// Restrict the live listing to enrichment-known models (drops non-chat
/// listing noise on polluted providers). Never adds models.
pub fn restrict_to_enriched(self) -> bool {
self.spec().restrict_to_enriched
}
/// Prefix to strip from live-listing model ids before filter/enrich/key
/// (e.g. Google's `models/`). `None` = use the id verbatim.
pub fn strip_listing_id_prefix(self) -> Option<&'static str> {
self.spec().strip_listing_id_prefix
}
/// Path (relative to base) for API-key validation at login. Defaults to
/// `/models`; a platform whose listing is public (OpenRouter) overrides
/// it with an auth-requiring endpoint so a bad key can't false-accept.
pub fn key_validation_path(self) -> &'static str {
self.spec().key_validation_path.unwrap_or("/models")
}
/// Model-listing endpoint shape + headers.
pub fn listing(self) -> ListingDialect {
self.spec().listing
}
/// Key header style for listing/validation/inference requests.
pub fn key_header(self) -> PlatformKeyHeader {
self.spec().key_header
}
/// ChatCompletions body-adaptation dialect (shell maps to the sampler's
/// `ChatCompat`; meaningless for other backends).
pub fn chat_compat(self) -> PlatformChatCompat {
self.spec().chat_compat
}
/// True ONLY for `github-copilot`: its `/models` listing and every
/// `/chat/completions` inference request must carry the VS Code Copilot
/// editor-identity headers (User-Agent / Editor-Version /
/// Editor-Plugin-Version / Copilot-Integration-Id). The shell gates the
/// listing headers and the sampler gates the inference headers on this, so
/// every other ChatCompletions platform's request stays byte-identical.
pub fn sends_copilot_editor_headers(self) -> bool {
matches!(self, Self::GithubCopilot)
}
/// True ONLY for `openai-codex`: its `/responses` inference request must
/// carry the Codex identity headers (`chatgpt-account-id` from the bearer
/// JWT, `originator: codex_cli_rs`, `OpenAI-Beta: responses=experimental`, a
/// codex `User-Agent`). The sampler gates these on this predicate, so the
/// API-key `openai` Responses requests stay byte-identical.
pub fn sends_codex_responses_headers(self) -> bool {
matches!(self, Self::OpenaiCodex)
}
/// The compiled-in catalog for a platform that serves NO live `/models`
/// listing (openai-codex), or `None` when the catalog comes from the wire.
///
/// openai-codex's 4 models are HARDCODED (read from the official Codex CLI's
/// `models_cache.json`, the `visibility=="list"` AND `supported_in_api==true`
/// set) because OpenAI exposes no stable public models endpoint for the
/// ChatGPT Codex backend. Each entry carries context window + per-model
/// selectable reasoning efforts (incl. the codex-only `xhigh`/`max`
/// tiers), so the fetch path maps them through the SAME
/// `platform_wire_model_to_entry` output as a live listing — no new type.
pub fn hardcoded_catalog(self) -> Option<Vec<WireModel>> {
match self {
Self::OpenaiCodex => Some(openai_codex_wire_models()),
_ => None,
}
}
}
/// One hardcoded openai-codex model as a [`WireModel`] (context 272000, thinking
/// + image input, selectable efforts). `efforts` are the exact per-model
/// supported tiers; `default` is the model's default effort.
fn codex_wire_model(slug: &str, display_name: &str, efforts: &[&str], default: &str) -> WireModel {
WireModel {
id: slug.to_string(),
context_length: 272_000,
supports_reasoning: true,
supports_image_in: true,
supports_video_in: false,
display_name: Some(display_name.to_string()),
max_output_tokens: 0,
supports_thinking_type: None,
think_efforts: Some(WireThinkEfforts {
support: true,
valid_efforts: efforts.iter().map(|s| (*s).to_string()).collect(),
default_effort: Some(default.to_string()),
}),
}
}
/// The HARDCODED openai-codex catalog: exactly the 4 `visibility=="list"` AND
/// `supported_in_api==true` models from the Codex CLI model cache. The
/// `gpt-5.3-codex-spark` (supported_in_api=false → not served by /responses),
/// `gpt-5.4`, `gpt-5.4-mini`, and `codex-auto-review` (visibility="hide") models
/// are deliberately EXCLUDED — they would list-but-not-work or are not
/// user-facing (fail-fast: never advertise a model the backend rejects).
fn openai_codex_wire_models() -> Vec<WireModel> {
vec![
codex_wire_model(
"gpt-5.6-sol",
"GPT-5.6-Sol",
&["low", "medium", "high", "xhigh", "max"],
"low",
),
codex_wire_model(
"gpt-5.6-terra",
"GPT-5.6-Terra",
&["low", "medium", "high", "xhigh", "max"],
"medium",
),
codex_wire_model(
"gpt-5.6-luna",
"GPT-5.6-Luna",
&["low", "medium", "high", "xhigh", "max"],
"medium",
),
codex_wire_model(
"gpt-5.5",
"GPT-5.5",
&["low", "medium", "high", "xhigh"],
"medium",
),
]
}
/// Split a managed catalog key `{platform_id}/{model_id}` back into its
/// platform and bare model id. `None` when the key carries no known platform
/// prefix (e.g. a user-defined `[model.*]` entry).
pub fn parse_managed_model_key(key: &str) -> Option<(PlatformId, &str)> {
let (platform, model_id) = key.split_once('/')?;
let platform = PlatformId::parse(platform)?;
if model_id.is_empty() {
return None;
}
Some((platform, model_id))
}
// Wire contract + capability derivation (PRD F4)
/// Model capabilities derived from the `/models` listing
/// (port of kimi-cli `ModelCapability` + `ModelInfo.capabilities`).
#[derive(
Debug, Clone, Copy, PartialEq, Eq, Hash, PartialOrd, Ord, serde::Serialize, serde::Deserialize,
)]
#[serde(rename_all = "snake_case")]
pub enum ModelCapability {
/// Supports reasoning ("thinking" mode toggleable on/off).
Thinking,
/// Thinking cannot be disabled (id contains "thinking").
AlwaysThinking,
ImageIn,
VideoIn,
}
impl ModelCapability {
pub fn as_str(self) -> &'static str {
match self {
Self::Thinking => "thinking",
Self::AlwaysThinking => "always_thinking",
Self::ImageIn => "image_in",
Self::VideoIn => "video_in",
}
}
}
impl std::fmt::Display for ModelCapability {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.write_str(self.as_str())
}
}
/// One entry of the `GET {base}/models` response `data` array (PRD F4).
#[derive(Debug, Clone, serde::Deserialize)]
pub struct WireModel {
pub id: String,
#[serde(default)]
pub context_length: u64,
#[serde(default)]
pub supports_reasoning: bool,
#[serde(default)]
pub supports_image_in: bool,
#[serde(default)]
pub supports_video_in: bool,
#[serde(default)]
pub display_name: Option<String>,
/// Max output tokens (`max_tokens` on the Anthropic listing; absent on
/// the Kimi/OpenAI-shape wires). 0 = unserved → enrichment may fill.
#[serde(default)]
pub max_output_tokens: u64,
/// `"only"` marks always-thinking models (thinking cannot be disabled).
/// Verified against the live `api.kimi.com/coding/v1/models` response.
#[serde(default)]
pub supports_thinking_type: Option<String>,
/// Selectable thinking-effort levels, present only on models that offer
/// them (e.g. K3). Verified against the live `/models` response.
#[serde(default)]
pub think_efforts: Option<WireThinkEfforts>,
}
/// The `think_efforts` object of a `/models` entry. Live wire shape
/// (api.kimi.com, 2026-07):
/// `{"support": true, "valid_efforts": ["low", "high", "max"],
/// "default_effort": "max"}`.
#[derive(Debug, Clone, Default, serde::Deserialize)]
pub struct WireThinkEfforts {
#[serde(default)]
pub support: bool,
#[serde(default)]
pub valid_efforts: Vec<String>,
#[serde(default)]
pub default_effort: Option<String>,
}
/// `GET {base}/models` response envelope.
#[derive(Debug, Clone, serde::Deserialize)]
pub struct WireModelsResponse {
pub data: Vec<WireModel>,
}
/// Parse an OpenAI-shape `/models` listing, tolerant of BOTH the standard
/// envelope `{object:"list", data:[...]}` and a bare top-level array `[...]`
/// (Together AI serves the bare-array form).
pub fn parse_openai_listing(json: &str) -> Result<Vec<WireModel>, serde_json::Error> {
// Sniff the top-level shape so a malformed body yields the diagnostic for
// the shape it actually is (a broken bare-array element reports the
// element error, not a misleading "expected the envelope object").
if json.trim_start().starts_with('[') {
serde_json::from_str::<Vec<WireModel>>(json)
} else {
Ok(serde_json::from_str::<WireModelsResponse>(json)?.data)
}
}
// Anthropic listing adapter (ListingDialect::Anthropic)
#[derive(serde::Deserialize)]
struct AnthropicListing {
/// No default: a 200 body without `data` is a contract violation and
/// must error like the OpenAI-shape branch, not yield an empty catalog.
data: Vec<AnthropicModel>,
#[serde(default)]
has_more: bool,
}
#[derive(serde::Deserialize, Default)]
#[serde(default)]
struct AnthropicModel {
id: String,
display_name: Option<String>,
max_input_tokens: u64,
/// The model's output cap (Anthropic REQUIRES `max_tokens` on
/// /v1/messages and 400s when it exceeds this).
max_tokens: u64,
capabilities: AnthropicCapabilities,
}
#[derive(serde::Deserialize, Default)]
#[serde(default)]
struct AnthropicCapabilities {
effort: AnthropicEffort,
thinking: AnthropicSupported,
image_input: AnthropicSupported,
}
#[derive(serde::Deserialize, Default)]
#[serde(default)]
struct AnthropicEffort {
supported: bool,
low: AnthropicSupported,
medium: AnthropicSupported,
high: AnthropicSupported,
xhigh: AnthropicSupported,
max: AnthropicSupported,
}
#[derive(serde::Deserialize, Default)]
#[serde(default)]
struct AnthropicSupported {
supported: bool,
}
/// Parse the 2026 Anthropic `GET /v1/models` response into the F4
/// [`WireModel`] shape: `max_input_tokens` → context (0 stays 0 so
/// enrichment can fill it), `capabilities.thinking/image_input` → flags,
/// `capabilities.effort.{level}.supported` → `think_efforts` in canonical
/// order. `has_more: true` (impossible under `?limit=1000` for Anthropic's
/// catalog size) warns rather than silently truncating.
pub fn parse_anthropic_listing(json: &str) -> Result<Vec<WireModel>, serde_json::Error> {
// Sniff the top-level shape. Anthropic itself serves the {data:[...]}
// envelope, but Anthropic-COMPATIBLE surfaces (e.g. MiniMax's /anthropic
// endpoint) may serve a bare array — accept both so such a provider isn't
// silently emptied (mirrors `parse_openai_listing`'s tolerance).
let data: Vec<AnthropicModel> = if json.trim_start().starts_with('[') {
serde_json::from_str::<Vec<AnthropicModel>>(json)?
} else {
let listing: AnthropicListing = serde_json::from_str(json)?;
if listing.has_more {
tracing::warn!(
fetched = listing.data.len(),
"anthropic /models reports more pages beyond limit=1000; \
listing may be incomplete"
);
}
listing.data
};
Ok(data
.into_iter()
.filter(|m| {
let keep = !m.id.is_empty();
if !keep {
tracing::warn!("anthropic listing entry without id; dropping");
}
keep
})
.map(|m| {
let e = &m.capabilities.effort;
let valid_efforts: Vec<String> = [
("low", e.low.supported),
("medium", e.medium.supported),
("high", e.high.supported),
("xhigh", e.xhigh.supported),
("max", e.max.supported),
]
.into_iter()
.filter(|(_, supported)| *supported)
.map(|(level, _)| level.to_string())
.collect();
// The wire has no default marker; the provider's implicit
// default applies until the user picks a level. An explicit
// `supported: false` becomes a DECLINE sentinel (support=false)
// — distinguishable from "wire silent", so enrichment can never
// inject a menu the server rejects (pre-4.6 models 400 on
// adaptive thinking).
let think_efforts = if e.supported && !valid_efforts.is_empty() {
Some(WireThinkEfforts {
support: true,
valid_efforts,
default_effort: None,
})
} else {
Some(WireThinkEfforts {
support: false,
valid_efforts: Vec::new(),
default_effort: None,
})
};
WireModel {
id: m.id,
context_length: m.max_input_tokens,
supports_reasoning: m.capabilities.thinking.supported,
supports_image_in: m.capabilities.image_input.supported,
supports_video_in: false,
display_name: m.display_name,
max_output_tokens: m.max_tokens,
supports_thinking_type: None,
think_efforts,
}
})
.collect())
}
impl WireModel {
/// Capability derivation ported verbatim from kimi-cli
/// `auth/platforms.py::ModelInfo.capabilities`:
/// - `supports_reasoning` → thinking
/// - `"thinking"` in id → thinking + always_thinking
/// - `supports_image_in` → image_in; `supports_video_in` → video_in
/// - id starts with `kimi-k2` → thinking + image_in + video_in
///
/// On top of that, the live wire's `supports_thinking_type: "only"`
/// marks a model whose thinking cannot be disabled → always_thinking.
///
/// Returned sorted + deduplicated ([`ModelCapability`]'s `Ord`).
pub fn capabilities(&self) -> Vec<ModelCapability> {
let mut caps = derive_capabilities(
&self.id,
self.supports_reasoning,
self.supports_image_in,
self.supports_video_in,
);
if self.supports_thinking_type.as_deref() == Some("only") {
for cap in [ModelCapability::Thinking, ModelCapability::AlwaysThinking] {
if !caps.contains(&cap) {
caps.push(cap);
}
}
caps.sort();
}
caps
}
}
/// See [`WireModel::capabilities`]; split out so fallback/bundled entries can
/// run the same derivation from an id alone.
pub fn derive_capabilities(
id: &str,
supports_reasoning: bool,
supports_image_in: bool,
supports_video_in: bool,
) -> Vec<ModelCapability> {
let id_lower = id.to_lowercase();
let mut caps = std::collections::BTreeSet::new();
if supports_reasoning {
caps.insert(ModelCapability::Thinking);
}
if id_lower.contains("thinking") {
caps.insert(ModelCapability::Thinking);
caps.insert(ModelCapability::AlwaysThinking);
}
if supports_image_in {
caps.insert(ModelCapability::ImageIn);
}
if supports_video_in {
caps.insert(ModelCapability::VideoIn);
}
if id_lower.starts_with("kimi-k2") {
caps.insert(ModelCapability::Thinking);
caps.insert(ModelCapability::ImageIn);
caps.insert(ModelCapability::VideoIn);
}
caps.into_iter().collect()
}
/// Whether thinking should default ON for a model with these capabilities
/// (PRD F4: `thinking` or `always_thinking` present).
pub fn default_thinking_enabled(capabilities: &[ModelCapability]) -> bool {
capabilities.iter().any(|c| {
matches!(
c,
ModelCapability::Thinking | ModelCapability::AlwaysThinking
)
})
}
/// Apply a platform's `allowed_model_prefixes` filter to a `/models` listing
/// (kimi-cli `list_models`). No-op for platforms without a filter.
pub fn filter_allowed_models(platform: PlatformId, models: Vec<WireModel>) -> Vec<WireModel> {
let Some(prefixes) = platform.allowed_model_prefixes() else {
return models;
};
models
.into_iter()
.filter(|m| prefixes.iter().any(|p| m.id.starts_with(p)))
.collect()
}
// GitHub Copilot listing adapter (github-copilot)
/// A Copilot model id that routes to the anthropic-messages wire in Pi
/// (`/^claude-(haiku|sonnet|opus)-[45]([.\-]|$)/`) — EXCLUDED from Kigi's
/// ChatCompletions catalog. Matches `claude-{haiku,sonnet,opus}-` followed by a
/// single `4`/`5` and then `.`, `-`, or end (so `claude-fable-5` is NOT a match
/// and stays served; `claude-sonnet-45` is not a real family and does not match).
fn is_copilot_messages_claude(id: &str) -> bool {
for family in ["claude-haiku-", "claude-sonnet-", "claude-opus-"] {
let Some(rest) = id.strip_prefix(family) else {
continue;
};
let mut chars = rest.chars();
if matches!(chars.next(), Some('4') | Some('5'))
&& matches!(chars.next(), None | Some('.') | Some('-'))
{
return true;
}
}
false
}
/// A Copilot model id served ONLY through the `/responses` endpoint in Pi
/// (`gpt-5*` / `oswe*` / `mai-*`) — EXCLUDED from Kigi's ChatCompletions catalog.
fn is_copilot_responses_only(id: &str) -> bool {
id.starts_with("gpt-5") || id.starts_with("oswe") || id.starts_with("mai-")
}
/// Whether a Copilot model id is served by the openai-completions wire — the
/// ONLY wire Kigi's github-copilot platform speaks. Excludes the claude-4.x/5.x
/// (messages) and gpt-5/oswe/mai- (responses-only) ids that would fail at
/// inference on `/chat/completions` (documented per-model-routing limitation).
pub fn is_copilot_completions_served(id: &str) -> bool {
!is_copilot_messages_claude(id) && !is_copilot_responses_only(id)
}
#[derive(serde::Deserialize)]
struct CopilotListing {
/// No default: a 200 body without `data` is a contract violation and must
/// error like the OpenAI-shape branch, not yield an empty catalog.
data: Vec<CopilotListingModel>,
}
#[derive(serde::Deserialize)]
struct CopilotListingModel {
id: String,
#[serde(default)]
name: Option<String>,
#[serde(default)]
model_picker_enabled: bool,
#[serde(default)]
policy: Option<CopilotPolicy>,
#[serde(default)]
capabilities: Option<CopilotCapabilities>,
}
#[derive(serde::Deserialize)]
struct CopilotPolicy {
#[serde(default)]
state: Option<String>,
}
#[derive(serde::Deserialize)]
struct CopilotCapabilities {
#[serde(default)]
supports: Option<CopilotSupports>,
}
#[derive(serde::Deserialize)]
struct CopilotSupports {
/// Absent means "not declined" → allowed (Pi: `supports.tool_calls !==
/// false`). Only an explicit `false` drops the model.
#[serde(default)]
tool_calls: Option<bool>,
}
/// Parse the GitHub Copilot `GET {base}/models` response and apply BOTH filters
/// (a Copilot quirk, gated on the platform):
/// 1. availability — keep iff `model_picker_enabled == true` AND
/// `policy.state != "disabled"` AND `capabilities.supports.tool_calls !=
/// false` (Pi `isSelectableCopilotModel`);
/// 2. wire-compat — keep iff the id is openai-completions-served (drops the
/// claude-4.x/5.x messages models and the gpt-5/oswe/mai- responses-only
/// models, which Kigi's single-wire platform cannot route).
///
/// Metadata (context window, thinking) is NOT served here — enrichment from
/// models.dev "github-copilot" fills it downstream.
pub fn parse_github_copilot_listing(json: &str) -> Result<Vec<WireModel>, serde_json::Error> {
let listing: CopilotListing = serde_json::from_str(json)?;
let kept = listing
.data
.into_iter()
.filter(|m| {
let picker_enabled = m.model_picker_enabled;
let policy_ok = m
.policy
.as_ref()
.and_then(|p| p.state.as_deref())
.is_none_or(|state| state != "disabled");
let tool_calls_ok = m
.capabilities
.as_ref()
.and_then(|c| c.supports.as_ref())
.and_then(|s| s.tool_calls)
!= Some(false);
picker_enabled && policy_ok && tool_calls_ok && is_copilot_completions_served(&m.id)
})
.map(|m| WireModel {
id: m.id,
context_length: 0,
supports_reasoning: false,
supports_image_in: false,
supports_video_in: false,
display_name: m.name,
max_output_tokens: 0,
supports_thinking_type: None,
think_efforts: None,
})
.collect();
Ok(kept)
}
// Bundled offline fallback catalog
/// The raw JSON, embedded at compile time. OFFLINE LAST RESORT: consulted only
/// when the live `/models` sync fails and no disk cache is usable.
///
/// Sources for every id (do not add ids that cannot be sourced):
/// - `kimi-for-coding`: kimi-cli `src/kimi_cli/llm.py` (`model_display_name`,
/// `derive_model_capabilities`) — the Kimi Code subscription coding model.
/// Its capabilities {thinking, image_in, video_in} come from
/// `derive_model_capabilities` in the same file.
/// - `kimi-k2-turbo-preview` / `kimi-k2-thinking-turbo`: kimi-cli
/// `tests/core/test_create_llm.py` (`_make_kimi_plain_model`,
/// `_make_kimi_thinking_model`) — Moonshot open-platform models. Their
/// capabilities follow the `auth/platforms.py` derivation rules
/// ([`derive_capabilities`]).
/// - context_window 262144: the canonical Kimi context size used by kimi-cli's
/// own budget tests (`tests/core/test_create_llm.py`).
///
/// Re-exported through the `kigi_shell::models` facade and consumed by
/// `agent::config`, so it must be `pub`.
pub const DEFAULT_MODELS_JSON: &str = include_str!("../default_models.json");
#[derive(serde::Deserialize)]
struct DefaultModels {
default: String,
/// Falls back to `default` if not specified in JSON.
image_description: Option<String>,
/// Falls back to `default` if not specified in JSON.
session_summary: Option<String>,
models: Vec<DefaultModelEntry>,
}
#[derive(serde::Deserialize)]
struct DefaultModelEntry {
model: String,
}
static DEFAULTS: LazyLock<DefaultModels> = LazyLock::new(|| {
let defaults: DefaultModels = serde_json::from_str(DEFAULT_MODELS_JSON)
.expect("default_models.json: invalid JSON or missing 'default' field");
// Baked-in JSON — a mismatch here is a developer error, not a runtime condition.
let model_ids: Vec<&str> = defaults.models.iter().map(|m| m.model.as_str()).collect();
assert!(
model_ids.contains(&defaults.default.as_str()),
"default_models.json: 'default' is '{}' but 'models' array only has {model_ids:?}",
defaults.default,
);
defaults
});
/// Primary model for coding tasks and general fallback.
pub fn default_model() -> &'static str {
&DEFAULTS.default
}
/// Model for image describe. Falls back to default model.
pub fn default_image_description_model() -> &'static str {
DEFAULTS
.image_description
.as_deref()
.unwrap_or(&DEFAULTS.default)
}
/// Model for session title generation. Falls back to default model.
pub fn default_session_summary_model() -> &'static str {
DEFAULTS
.session_summary
.as_deref()
.unwrap_or(&DEFAULTS.default)
}
#[cfg(test)]
mod tests {
use super::*;
/// Mirror of the live `api.kimi.com/coding/v1/models` K3 entry
/// (fetched 2026-07-17): `supports_thinking_type: "only"` plus a
/// `think_efforts` block with low/high/max and a max default.
#[test]
fn wire_model_parses_live_k3_think_efforts() {
let json = serde_json::json!({
"id": "k3",
"created": 1_761_264_000,
"object": "model",
"display_name": "K3",
"type": "model",
"context_length": 1_048_576,
"supports_reasoning": true,
"supports_image_in": true,
"supports_video_in": true,
"supports_thinking_type": "only",
"think_efforts": {
"support": true,
"valid_efforts": ["low", "high", "max"],
"default_effort": "max"
}
});
let wire: WireModel = serde_json::from_value(json).unwrap();
let efforts = wire.think_efforts.as_ref().unwrap();
assert!(efforts.support);
assert_eq!(efforts.valid_efforts, ["low", "high", "max"]);
assert_eq!(efforts.default_effort.as_deref(), Some("max"));
// "only" thinking type forces always_thinking on top of the
// supports_reasoning-derived thinking capability.
let caps = wire.capabilities();
assert!(caps.contains(&ModelCapability::Thinking));
assert!(caps.contains(&ModelCapability::AlwaysThinking));
}
/// The K2.7 entries carry `supports_thinking_type: "only"` but no
/// `think_efforts` — always-thinking without selectable levels.
#[test]
fn wire_model_without_think_efforts_still_always_thinking() {
let json = serde_json::json!({
"id": "kimi-for-coding",
"context_length": 262_144,
"supports_reasoning": true,
"supports_image_in": true,
"supports_video_in": true,
"supports_thinking_type": "only"
});
let wire: WireModel = serde_json::from_value(json).unwrap();
assert!(wire.think_efforts.is_none());
let caps = wire.capabilities();
assert!(caps.contains(&ModelCapability::AlwaysThinking));
// Sorted + deduplicated invariant holds after the "only" injection.
let mut sorted = caps.clone();
sorted.sort();
sorted.dedup();
assert_eq!(caps, sorted);
}
/// The Anthropic listing adapter maps the documented 2026 response shape
/// (platform.claude.com/docs/en/api/models-list) onto WireModel: effort
/// capability levels become think_efforts in canonical order, a zero
/// max_input_tokens stays zero (enrichment fills it), unknown fields
/// tolerated, effort.supported=false yields no menu.
#[test]
fn anthropic_listing_maps_documented_shape() {
let json = serde_json::json!({
"data": [
{
"id": "claude-opus-4-6",
"display_name": "Claude Opus 4.6",
"created_at": "2026-02-04T00:00:00Z",
"type": "model",
"max_input_tokens": 1_000_000,
"max_tokens": 128_000,
"created_at_is_ignored": true,
"capabilities": {
"batch": { "supported": true },
"effort": {
"supported": true,
"low": { "supported": true },
"medium": { "supported": true },
"high": { "supported": true },
"xhigh": { "supported": true },
"max": { "supported": true }
},
"thinking": {
"supported": true,
"types": {
"adaptive": { "supported": true },
"enabled": { "supported": true }
}
},
"image_input": { "supported": true },
"structured_outputs": { "supported": true }
}
},
{
"id": "claude-legacy",
"max_input_tokens": 0,
"capabilities": {
"effort": { "supported": false },
"thinking": { "supported": false },
"image_input": { "supported": false }
}
}
],
"first_id": "claude-opus-4-6",
"last_id": "claude-legacy",
"has_more": false
})
.to_string();
let models = parse_anthropic_listing(&json).expect("documented shape parses");
assert_eq!(models.len(), 2);
let opus = &models[0];
assert_eq!(opus.id, "claude-opus-4-6");
assert_eq!(opus.context_length, 1_000_000);
assert_eq!(
opus.max_output_tokens, 128_000,
"wire max_tokens is the output cap (Anthropic 400s above it)"
);
assert!(opus.supports_reasoning && opus.supports_image_in);
assert_eq!(opus.display_name.as_deref(), Some("Claude Opus 4.6"));
let efforts = opus.think_efforts.as_ref().expect("effort menu");
assert_eq!(
efforts.valid_efforts,
["low", "medium", "high", "xhigh", "max"],
"levels in canonical order from per-level supported flags"
);
assert_eq!(efforts.default_effort, None);
let legacy = &models[1];
assert_eq!(legacy.context_length, 0, "zero stays zero for enrichment");
let decline = legacy
.think_efforts
.as_ref()
.expect("explicit wire decline is a sentinel, not absence");
assert!(
!decline.support && decline.valid_efforts.is_empty(),
"effort.supported=false must block enrichment menu injection"
);
assert!(!legacy.supports_reasoning);
// A 200 body without `data` is a contract violation, not an empty
// catalog; entries without an id are dropped with a warning.
assert!(parse_anthropic_listing("{}").is_err());
let ghosts = serde_json::json!({ "data": [ {}, { "id": "real" } ] }).to_string();
let models = parse_anthropic_listing(&ghosts).unwrap();
assert_eq!(
models.iter().map(|m| m.id.as_str()).collect::<Vec<_>>(),
vec!["real"]
);
}
/// Anthropic-COMPATIBLE surfaces (e.g. MiniMax's /anthropic endpoint) may
/// serve a bare array instead of the `{data:[...]}` envelope; the parser
/// accepts both so such a provider isn't silently emptied. A bare object
/// (not an array, no `data`) still errors.
#[test]
fn anthropic_listing_accepts_envelope_and_bare_array() {
let envelope =
serde_json::json!({ "data": [ { "id": "MiniMax-M2.5", "max_input_tokens": 204_800 } ] })
.to_string();
let bare = serde_json::json!([ { "id": "MiniMax-M2.5", "max_input_tokens": 204_800 } ])
.to_string();
for json in [&envelope, &bare] {
let models = parse_anthropic_listing(json).expect("both shapes parse");
assert_eq!(models.len(), 1);
assert_eq!(models[0].id, "MiniMax-M2.5");
assert_eq!(models[0].context_length, 204_800);
}
// A bare object without `data` is still a contract violation.
assert!(parse_anthropic_listing(r#"{"foo":1}"#).is_err());
}
/// The OpenAI listing parser accepts both the standard envelope and a
/// bare top-level array (Together AI serves the bare form).
#[test]
fn openai_listing_parse_accepts_envelope_and_bare_array() {
let envelope = r#"{"object":"list","data":[
{"id":"a","context_length":1000},{"id":"b"}]}"#;
let bare = r#"[{"id":"a","context_length":1000},{"id":"b"}]"#;
for (label, json) in [("envelope", envelope), ("bare array", bare)] {
let models =
parse_openai_listing(json).unwrap_or_else(|e| panic!("{label} must parse: {e}"));
assert_eq!(
models.iter().map(|m| m.id.as_str()).collect::<Vec<_>>(),
vec!["a", "b"],
"{label}"
);
assert_eq!(models[0].context_length, 1000);
}
// Neither shape → error (not a silent empty list).
assert!(parse_openai_listing("\"not a list\"").is_err());
assert!(parse_openai_listing("{").is_err());
}
/// Providers whose `/models` listing is public must validate keys
/// against an auth-requiring endpoint; every other platform validates
/// against the default `/models`.
#[test]
fn public_listing_providers_override_the_validation_path() {
assert_eq!(PlatformId::OpenRouter.key_validation_path(), "/key");
assert_eq!(PlatformId::Vercel.key_validation_path(), "/credits");
let overrides = [PlatformId::OpenRouter, PlatformId::Vercel];
for p in PlatformId::ALL {
if !overrides.contains(&p) {
assert_eq!(
p.key_validation_path(),
"/models",
"{} must validate against /models",
p.as_str()
);
}
}
}
/// Google's compat listing returns `models/`-prefixed ids; the spec
/// must declare the strip so they canonicalize to the bare snapshot
/// form. Every other platform uses ids verbatim.
#[test]
fn only_google_strips_a_listing_id_prefix() {
assert_eq!(
PlatformId::Google.strip_listing_id_prefix(),
Some("models/")
);
for p in PlatformId::ALL {
if p != PlatformId::Google {
assert_eq!(
p.strip_listing_id_prefix(),
None,
"{} must use listing ids verbatim",
p.as_str()
);
}
}
}
#[test]
fn platform_ids_round_trip() {
for p in PlatformId::ALL {
assert_eq!(PlatformId::parse(p.as_str()), Some(p));
}
assert_eq!(PlatformId::parse("not-a-platform"), None);
}
/// xai-grok is the second refreshable-OAuth platform: it carries a generic
/// device-code `OAuthConfig` (Kimi Code carries `None`), reuses the xai
/// wire (models.dev "xai", restrict-to-enriched, Bearer, ChatCompletions),
/// and keys its models under `xai-grok/` — distinct from the API-key `xai`.
#[test]
fn xai_grok_is_a_generic_oauth_platform() {
let g = PlatformId::XaiGrok;
assert_eq!(g.as_str(), "xai-grok");
assert!(g.uses_oauth());
// Kimi Code is uses_oauth yet carries no generic config (bespoke flow).
assert!(PlatformId::KimiCode.uses_oauth());
assert_eq!(PlatformId::KimiCode.oauth(), None);
let cfg = g
.oauth()
.expect("xai-grok carries a device-code OAuthConfig");
assert_eq!(cfg, &XAI_OAUTH_CONFIG);
assert_eq!(cfg.client_id, "b1a00492-073a-47ea-816f-4c329264a828");
assert_eq!(cfg.auth_host, "https://auth.x.ai");
assert_eq!(cfg.device_path, "/oauth2/device/code");
assert_eq!(cfg.token_path, "/oauth2/token");
assert_eq!(
cfg.scope,
"openid profile email offline_access grok-cli:access api:access"
);
assert_eq!(cfg.scope_key, "oauth/xai");
assert_eq!(cfg.extra_device_field, Some(("referrer", "kigi")));
// Scope-key lookup resolves the config (drives the generic refresher).
assert_eq!(
oauth_config_for_scope_key("oauth/xai"),
Some(&XAI_OAUTH_CONFIG)
);
assert_eq!(oauth_config_for_scope_key("oauth/kimi-code"), None);
// Reuses the xai wire, its OWN base env, and the models.dev "xai" keys.
assert_eq!(g.models_dev_id(), Some("xai"));
assert!(g.restrict_to_enriched());
assert_eq!(g.key_header(), PlatformKeyHeader::Bearer);
assert_eq!(g.wire_api(), PlatformWireApi::ChatCompletions);
assert_eq!(g.api_key_env_names(), &[] as &[&str]);
assert_eq!(g.managed_model_key("grok-4.5"), "xai-grok/grok-4.5");
let _guard = kigi_env::EnvVarGuard::set(XAI_GROK_BASE_URL_ENV, "https://mock.grok/v1");
assert_eq!(g.base_url(), "https://mock.grok/v1");
}
/// claude-pro-max is the first PKCE-localhost OAuth platform: it carries a
/// PKCE `OAuthConfig` (authorize host ≠ token host, JSON token body), reuses
/// the Anthropic Messages + listing wire with a Bearer key header (OAuth,
/// NOT x-api-key), enriches from models.dev "anthropic", and keys its models
/// under `claude-pro-max/`.
#[test]
fn claude_pro_max_is_a_pkce_oauth_platform() {
let c = PlatformId::ClaudeProMax;
assert_eq!(c.as_str(), "claude-pro-max");
assert!(c.uses_oauth());
let cfg = c
.oauth()
.expect("claude-pro-max carries a PKCE OAuthConfig");
assert_eq!(cfg, &CLAUDE_OAUTH_CONFIG);
assert_eq!(cfg.client_id, "9d1c250a-e61b-44d9-88ed-5944d1962f5e");
// Authorize host ≠ token host (the distinguishing PKCE trait).
assert_eq!(cfg.auth_host, "https://claude.ai");
assert_eq!(cfg.device_path, "/oauth/authorize");
assert_eq!(cfg.token_host, "https://platform.claude.com");
assert_eq!(cfg.token_path, "/v1/oauth/token");
assert_eq!(
cfg.scope,
"org:create_api_key user:profile user:inference \
user:sessions:claude_code user:mcp_servers user:file_upload"
);
assert_eq!(cfg.scope_key, "oauth/claude-pro-max");
assert_eq!(cfg.extra_device_field, None);
assert_eq!(
cfg.flow,
OAuthFlow::PkceLocalhost {
redirect_port: 53692,
redirect_path: "/callback",
}
);
assert_eq!(cfg.token_body, OAuthTokenBody::Json);
// xai stays the device-code / form contract — unaffected.
assert_eq!(XAI_OAUTH_CONFIG.flow, OAuthFlow::DeviceCode);
assert_eq!(XAI_OAUTH_CONFIG.token_body, OAuthTokenBody::Form);
assert_eq!(XAI_OAUTH_CONFIG.token_host, "https://auth.x.ai");
// Scope-key lookup resolves the config (drives the generic refresher).
assert_eq!(
oauth_config_for_scope_key("oauth/claude-pro-max"),
Some(&CLAUDE_OAUTH_CONFIG)
);
// Anthropic Messages + listing wire, reached with a Bearer OAuth token.
assert_eq!(c.models_dev_id(), Some("anthropic"));
assert!(!c.restrict_to_enriched());
assert_eq!(c.key_header(), PlatformKeyHeader::Bearer);
assert_eq!(c.wire_api(), PlatformWireApi::Messages);
assert_eq!(c.listing(), ListingDialect::Anthropic);
assert_eq!(c.api_key_env_names(), &[] as &[&str]);
assert_eq!(
c.managed_model_key("claude-opus-4-8"),
"claude-pro-max/claude-opus-4-8"
);
let _guard =
kigi_env::EnvVarGuard::set(CLAUDE_OAUTH_BASE_URL_ENV, "https://mock.claude/v1");
assert_eq!(c.base_url(), "https://mock.claude/v1");
}
/// github-copilot is the two-stage device-code OAuth platform: it carries a
/// `GithubDeviceCopilot`/`GithubCopilotExchange` `OAuthConfig` with a
/// copilot-token exchange endpoint (Stage 2), speaks the ChatCompletions
/// wire with a Bearer copilot token + the editor-headers gate, enriches from
/// models.dev "github-copilot", and keys its models under `github-copilot/`.
#[test]
fn github_copilot_is_a_two_stage_oauth_platform() {
let g = PlatformId::GithubCopilot;
assert_eq!(g.as_str(), "github-copilot");
assert!(g.uses_oauth());
assert!(
g.sends_copilot_editor_headers(),
"github-copilot must gate the editor-identity headers"
);
// Every OTHER platform must NOT send the editor headers (regression).
for other in PlatformId::ALL {
if other != PlatformId::GithubCopilot {
assert!(
!other.sends_copilot_editor_headers(),
"{} must not send the copilot editor headers",
other.as_str()
);
}
}
let cfg = g
.oauth()
.expect("github-copilot carries a two-stage OAuthConfig");
assert_eq!(cfg, &COPILOT_OAUTH_CONFIG);
assert_eq!(cfg.client_id, "Iv1.b507a08c87ecfe98");
assert_eq!(cfg.auth_host, "https://github.com");
assert_eq!(cfg.device_path, "/login/device/code");
assert_eq!(cfg.token_path, "/login/oauth/access_token");
assert_eq!(cfg.scope, "read:user");
assert_eq!(cfg.scope_key, "oauth/github-copilot");
assert_eq!(cfg.flow, OAuthFlow::GithubDeviceCopilot);
assert_eq!(cfg.token_body, OAuthTokenBody::GithubCopilotExchange);
assert_eq!(
cfg.copilot_exchange,
Some(("https://api.github.com", "/copilot_internal/v2/token")),
"the Stage-2 copilot-token exchange endpoint must be configured"
);
// xai/claude carry no copilot exchange (their refresh is a plain grant).
assert_eq!(XAI_OAUTH_CONFIG.copilot_exchange, None);
assert_eq!(CLAUDE_OAUTH_CONFIG.copilot_exchange, None);
// Scope-key lookup resolves the config (drives the generic refresher's
// Copilot re-mint dispatch).
assert_eq!(
oauth_config_for_scope_key("oauth/github-copilot"),
Some(&COPILOT_OAUTH_CONFIG)
);
// ChatCompletions wire, Bearer, models.dev "github-copilot", own base.
assert_eq!(g.models_dev_id(), Some("github-copilot"));
assert_eq!(g.wire_api(), PlatformWireApi::ChatCompletions);
assert_eq!(g.listing(), ListingDialect::OpenAi);
assert_eq!(g.key_header(), PlatformKeyHeader::Bearer);
assert_eq!(g.api_key_env_names(), &[] as &[&str]);
assert!(!g.restrict_to_enriched());
assert_eq!(g.managed_model_key("gpt-4.1"), "github-copilot/gpt-4.1");
let _guard = kigi_env::EnvVarGuard::set(COPILOT_BASE_URL_ENV, "https://mock.copilot");
assert_eq!(g.base_url(), "https://mock.copilot");
}
/// openai-codex is the ChatGPT/Codex PKCE-localhost OAuth platform: it
/// carries a `PkceLocalhost{1455, "/auth/callback"}` / FORM `OAuthConfig`
/// with the 3 authorize-extra params, speaks the Responses wire with a Bearer
/// OAuth token + the codex-headers gate, and is NOT models.dev-enriched (its
/// catalog is hardcoded).
#[test]
fn openai_codex_is_a_pkce_responses_oauth_platform() {
let c = PlatformId::OpenaiCodex;
assert_eq!(c.as_str(), "openai-codex");
assert!(c.uses_oauth());
assert!(
c.sends_codex_responses_headers(),
"openai-codex must gate the Codex identity headers"
);
// Every OTHER platform must NOT send the codex headers (regression).
for other in PlatformId::ALL {
if other != PlatformId::OpenaiCodex {
assert!(
!other.sends_codex_responses_headers(),
"{} must not send the codex responses headers",
other.as_str()
);
}
}
let cfg = c.oauth().expect("openai-codex carries a PKCE OAuthConfig");
assert_eq!(cfg, &CODEX_OAUTH_CONFIG);
assert_eq!(cfg.client_id, "app_EMoamEEZ73f0CkXaXp7hrann");
assert_eq!(cfg.auth_host, "https://auth.openai.com");
assert_eq!(cfg.token_host, "https://auth.openai.com");
assert_eq!(cfg.device_path, "/oauth/authorize");
assert_eq!(cfg.token_path, "/oauth/token");
assert_eq!(cfg.scope, "openid profile email offline_access");
assert_eq!(cfg.scope_key, "oauth/openai-codex");
assert_eq!(
cfg.flow,
OAuthFlow::PkceLocalhost {
redirect_port: 1455,
redirect_path: "/auth/callback",
}
);
// FORM token body (refresh routes through the generic device refresher).
assert_eq!(cfg.token_body, OAuthTokenBody::Form);
assert_eq!(cfg.copilot_exchange, None);
// The 3 authorize-only extra params; claude/xai/copilot carry none.
assert_eq!(
cfg.authorize_extra,
&[
("id_token_add_organizations", "true"),
("codex_cli_simplified_flow", "true"),
("originator", "codex_cli_rs"),
]
);
assert_eq!(CLAUDE_OAUTH_CONFIG.authorize_extra, &[] as &[(&str, &str)]);
assert_eq!(XAI_OAUTH_CONFIG.authorize_extra, &[] as &[(&str, &str)]);
assert_eq!(COPILOT_OAUTH_CONFIG.authorize_extra, &[] as &[(&str, &str)]);
// The chatgpt_account_id requirement is an EXPLICIT per-provider fact,
// never inferred from the token-body encoding — a future form-encoded
// PKCE provider must not inherit ChatGPT's account-id gate. Enforced at
// COMPILE time: adding a provider that flips this fails the build.
const {
assert!(CODEX_OAUTH_CONFIG.requires_chatgpt_account_id);
assert!(!CLAUDE_OAUTH_CONFIG.requires_chatgpt_account_id);
assert!(!XAI_OAUTH_CONFIG.requires_chatgpt_account_id);
assert!(!COPILOT_OAUTH_CONFIG.requires_chatgpt_account_id);
}
assert_eq!(
oauth_config_for_scope_key("oauth/openai-codex"),
Some(&CODEX_OAUTH_CONFIG)
);
// Responses wire, Bearer, hardcoded (no models.dev id), own base.
assert_eq!(c.models_dev_id(), None);
assert_eq!(c.wire_api(), PlatformWireApi::Responses);
assert_eq!(c.key_header(), PlatformKeyHeader::Bearer);
assert_eq!(c.api_key_env_names(), &[] as &[&str]);
assert_eq!(c.managed_model_key("gpt-5.5"), "openai-codex/gpt-5.5");
assert_eq!(
c.base_url(),
"https://chatgpt.com/backend-api/codex",
"default codex base carries the /codex tail (→ /codex/responses)"
);
let _guard = kigi_env::EnvVarGuard::set(CODEX_BASE_URL_ENV, "https://mock.codex/codex");
assert_eq!(c.base_url(), "https://mock.codex/codex");
}
/// The HARDCODED openai-codex catalog is exactly the 4 supported+listed
/// models, keyed by slug, ctx 272000, each exposing its exact supported
/// efforts (incl. the codex-only `xhigh`/`max` tiers). The
/// list-but-broken / hidden models are absent. Every other platform serves
/// NO hardcoded catalog (its models come from the live wire).
#[test]
fn openai_codex_hardcoded_catalog_is_the_four_supported_models() {
let catalog = PlatformId::OpenaiCodex
.hardcoded_catalog()
.expect("openai-codex serves a hardcoded catalog");
let ids: Vec<&str> = catalog.iter().map(|m| m.id.as_str()).collect();
assert_eq!(
ids,
vec!["gpt-5.6-sol", "gpt-5.6-terra", "gpt-5.6-luna", "gpt-5.5"],
"exactly the 4 visibility=list AND supported_in_api=true models"
);
// Excluded: list-but-not-served + hidden models never appear.
for absent in [
"gpt-5.3-codex-spark",
"gpt-5.4",
"gpt-5.4-mini",
"codex-auto-review",
] {
assert!(
!ids.contains(&absent),
"{absent} must be excluded from the hardcoded catalog"
);
}
for m in &catalog {
assert_eq!(m.context_length, 272_000, "{} ctx", m.id);
assert!(m.supports_reasoning && m.supports_image_in, "{} caps", m.id);
let caps = m.capabilities();
assert!(caps.contains(&ModelCapability::Thinking));
assert!(caps.contains(&ModelCapability::ImageIn));
}
// Per-model efforts (the crux of "their thinking method").
let efforts = |slug: &str| -> Vec<String> {
catalog
.iter()
.find(|m| m.id == slug)
.unwrap()
.think_efforts
.as_ref()
.unwrap()
.valid_efforts
.clone()
};
assert_eq!(
efforts("gpt-5.6-sol"),
["low", "medium", "high", "xhigh", "max"]
);
assert_eq!(
efforts("gpt-5.6-terra"),
["low", "medium", "high", "xhigh", "max"]
);
assert_eq!(
efforts("gpt-5.6-luna"),
["low", "medium", "high", "xhigh", "max"]
);
assert_eq!(efforts("gpt-5.5"), ["low", "medium", "high", "xhigh"]);
// Default efforts per the model table.
let default = |slug: &str| -> Option<String> {
catalog
.iter()
.find(|m| m.id == slug)
.unwrap()
.think_efforts
.as_ref()
.unwrap()
.default_effort
.clone()
};
assert_eq!(default("gpt-5.6-sol").as_deref(), Some("low"));
assert_eq!(default("gpt-5.6-terra").as_deref(), Some("medium"));
// Only openai-codex has a hardcoded catalog; every wire platform is None.
for p in PlatformId::ALL {
if p != PlatformId::OpenaiCodex {
assert!(
p.hardcoded_catalog().is_none(),
"{} must not carry a hardcoded catalog",
p.as_str()
);
}
}
}
/// The Copilot `/models` filter keeps ONLY openai-completions-served,
/// selectable, tool-calling models — dropping the claude-4.x/5.x (messages)
/// and gpt-5/oswe/mai- (responses-only) ids AND the disabled / picker-off /
/// tool-call-declining ones — with the display name carried through.
#[test]
fn copilot_listing_filters_by_availability_and_wire_compat() {
let body = serde_json::json!({ "data": [
// KEPT: completions-served, picker on, no policy, tool_calls absent.
{ "id": "gpt-4.1", "name": "GPT-4.1", "model_picker_enabled": true },
// KEPT: policy enabled + tool_calls true.
{ "id": "gemini-3-flash-preview", "model_picker_enabled": true,
"policy": {"state": "enabled"},
"capabilities": {"supports": {"tool_calls": true}} },
// KEPT: claude-fable-5 is NOT a messages-routed claude family.
{ "id": "claude-fable-5", "model_picker_enabled": true },
// DROPPED: claude-4.x → anthropic-messages wire.
{ "id": "claude-opus-4-8", "model_picker_enabled": true },
// DROPPED: claude-5.x → anthropic-messages wire.
{ "id": "claude-sonnet-5", "model_picker_enabled": true },
// DROPPED: gpt-5* → responses-only.
{ "id": "gpt-5.2", "model_picker_enabled": true },
// DROPPED: mai-* → responses-only.
{ "id": "mai-code-1", "model_picker_enabled": true },
// DROPPED: picker disabled.
{ "id": "gpt-4o", "model_picker_enabled": false },
// DROPPED: policy disabled.
{ "id": "kimi-k2.7-code", "model_picker_enabled": true,
"policy": {"state": "disabled"} },
// DROPPED: tool_calls explicitly false.
{ "id": "text-embed", "model_picker_enabled": true,
"capabilities": {"supports": {"tool_calls": false}} }
]})
.to_string();
let kept = parse_github_copilot_listing(&body).expect("valid listing");
let ids: Vec<&str> = kept.iter().map(|m| m.id.as_str()).collect();
assert_eq!(
ids,
vec!["gpt-4.1", "gemini-3-flash-preview", "claude-fable-5"],
"only completions-served, selectable, tool-calling models survive"
);
assert_eq!(
kept[0].display_name.as_deref(),
Some("GPT-4.1"),
"the wire display name must carry through"
);
// A 200 body without `data` is a contract violation (never empty catalog).
assert!(parse_github_copilot_listing("{}").is_err());
}
/// A variant missing from `ALL` compiles fine (`ALL`'s length is a plain
/// literal) but is silently unparseable and excluded from model sync.
/// The exhaustive match below fails compilation when a variant is added,
/// forcing this test — and with it the `ALL` entry — to be updated.
#[test]
fn all_covers_every_variant() {
fn ordinal(p: PlatformId) -> usize {
match p {
PlatformId::KimiCode => 0,
PlatformId::MoonshotCn => 1,
PlatformId::MoonshotAi => 2,
PlatformId::OpenAi => 3,
PlatformId::Anthropic => 4,
PlatformId::DeepSeek => 5,
PlatformId::Groq => 6,
PlatformId::Mistral => 7,
PlatformId::Fireworks => 8,
PlatformId::Google => 9,
PlatformId::OpenRouter => 10,
PlatformId::Together => 11,
PlatformId::Cerebras => 12,
PlatformId::Nvidia => 13,
PlatformId::Vercel => 14,
PlatformId::Xai => 15,
PlatformId::QwenTokenPlan => 16,
PlatformId::QwenTokenPlanCn => 17,
PlatformId::KimiCoding => 18,
PlatformId::Zai => 19,
PlatformId::ZaiCodingCn => 20,
PlatformId::Xiaomi => 21,
PlatformId::XiaomiTokenPlanCn => 22,
PlatformId::Minimax => 23,
PlatformId::MinimaxCn => 24,
PlatformId::XaiGrok => 25,
PlatformId::ClaudeProMax => 26,
PlatformId::GithubCopilot => 27,
PlatformId::OpenaiCodex => 28,
}
}
// update together with `ordinal`
const VARIANT_COUNT: usize = 29;
let mut seen: Vec<usize> = PlatformId::ALL.iter().map(|&p| ordinal(p)).collect();
seen.sort_unstable();
seen.dedup();
assert_eq!(
seen.len(),
PlatformId::ALL.len(),
"duplicate variant in ALL"
);
assert_eq!(
seen.len(),
VARIANT_COUNT,
"PlatformId::ALL must contain every variant"
);
}
/// Row-shape invariants the login UI and key resolution rely on:
/// API-key platforms carry a console host (paste-box copy) and at least
/// one key env var (missing-key error names it); OAuth channels carry
/// neither key envs nor a console host requirement.
#[test]
fn api_key_rows_carry_console_host_and_env_names() {
for p in PlatformId::ALL {
if p.uses_oauth() {
assert!(
p.api_key_env_names().is_empty(),
"{}: OAuth platforms take no key envs",
p.as_str()
);
} else {
assert!(
p.console_host().is_some(),
"{}: API-key platforms must name their console host",
p.as_str()
);
assert!(
!p.api_key_env_names().is_empty(),
"{}: API-key platforms must name at least one key env",
p.as_str()
);
assert!(
!p.vendor().is_empty(),
"{}: API-key platforms must set a vendor word",
p.as_str()
);
}
}
}
/// `parse` resolves by scanning spec rows, so duplicate ids would
/// silently shadow a platform. Pin uniqueness as rows are added.
#[test]
fn platform_spec_ids_unique() {
let mut ids: Vec<&str> = PlatformId::ALL.iter().map(|p| p.as_str()).collect();
ids.sort_unstable();
let before = ids.len();
ids.dedup();
assert_eq!(ids.len(), before, "duplicate platform spec id");
}
#[test]
fn platform_base_urls() {
assert_eq!(
PlatformId::MoonshotCn.base_url(),
"https://api.moonshot.cn/v1"
);
assert_eq!(
PlatformId::MoonshotAi.base_url(),
"https://api.moonshot.ai/v1"
);
// Subscription base honors the env override.
let _g = kigi_env::EnvVarGuard::set(kigi_env::CODE_BASE_URL_ENV, "https://mock.test/v1");
assert_eq!(PlatformId::KimiCode.base_url(), "https://mock.test/v1");
}
#[test]
fn managed_model_key_format_and_parse() {
let key = PlatformId::MoonshotCn.managed_model_key("kimi-k2-turbo-preview");
assert_eq!(key, "moonshot-cn/kimi-k2-turbo-preview");
assert_eq!(
parse_managed_model_key(&key),
Some((PlatformId::MoonshotCn, "kimi-k2-turbo-preview"))
);
assert_eq!(
parse_managed_model_key("kimi-code/kimi-for-coding"),
Some((PlatformId::KimiCode, "kimi-for-coding"))
);
// FIRST-slash split: provider-native slashed ids (11 of 15 groq
// models, e.g. openai/gpt-oss-120b) must survive the round trip.
assert_eq!(
parse_managed_model_key("groq/openai/gpt-oss-120b"),
Some((PlatformId::Groq, "openai/gpt-oss-120b"))
);
// No prefix / unknown platform / empty model id → None.
assert_eq!(parse_managed_model_key("kimi-for-coding"), None);
assert_eq!(parse_managed_model_key("not-a-platform/gpt"), None);
assert_eq!(parse_managed_model_key("moonshot-cn/"), None);
}
/// Capability derivation table ported from kimi-cli platforms.py.
#[test]
fn capability_derivation_table() {
use ModelCapability::*;
let cases: &[(&str, bool, bool, bool, &[ModelCapability])] = &[
// supports_reasoning only → thinking
("some-model", true, false, false, &[Thinking]),
// no flags, no name rules → empty
("some-model", false, false, false, &[]),
// "thinking" in id → thinking + always_thinking
(
"kimi-latest-thinking",
false,
false,
false,
&[Thinking, AlwaysThinking],
),
// image/video flags map directly
("some-model", false, true, true, &[ImageIn, VideoIn]),
// kimi-k2 prefix → thinking + image_in + video_in
(
"kimi-k2-turbo-preview",
false,
false,
false,
&[Thinking, ImageIn, VideoIn],
),
// kimi-k2 prefix + "thinking" in id → all four
(
"kimi-k2-thinking-turbo",
false,
false,
false,
&[Thinking, AlwaysThinking, ImageIn, VideoIn],
),
// Case-insensitive id rules (mirrors `.lower()` in platforms.py)
(
"Kimi-K2-Thinking",
false,
false,
false,
&[Thinking, AlwaysThinking, ImageIn, VideoIn],
),
];
for (id, reasoning, image, video, want) in cases {
let got = derive_capabilities(id, *reasoning, *image, *video);
assert_eq!(&got, want, "capabilities for {id}");
}
}
#[test]
fn default_thinking_from_capabilities() {
use ModelCapability::*;
assert!(default_thinking_enabled(&[Thinking]));
assert!(default_thinking_enabled(&[AlwaysThinking]));
assert!(default_thinking_enabled(&[Thinking, ImageIn]));
assert!(!default_thinking_enabled(&[ImageIn, VideoIn]));
assert!(!default_thinking_enabled(&[]));
}
#[test]
fn moonshot_prefix_filter_applies_only_to_open_platforms() {
let listing = vec![
WireModel {
id: "kimi-k2-turbo-preview".into(),
context_length: 262_144,
supports_reasoning: false,
supports_image_in: false,
supports_video_in: false,
display_name: None,
max_output_tokens: 0,
supports_thinking_type: None,
think_efforts: None,
},
WireModel {
id: "moonshot-v1-8k".into(),
context_length: 8_192,
supports_reasoning: false,
supports_image_in: false,
supports_video_in: false,
display_name: None,
max_output_tokens: 0,
supports_thinking_type: None,
think_efforts: None,
},
];
let filtered = filter_allowed_models(PlatformId::MoonshotCn, listing.clone());
assert_eq!(
filtered.iter().map(|m| m.id.as_str()).collect::<Vec<_>>(),
vec!["kimi-k2-turbo-preview"],
"moonshot listing must be filtered to the kimi-k prefix"
);
let unfiltered = filter_allowed_models(PlatformId::KimiCode, listing);
assert_eq!(unfiltered.len(), 2, "subscription listing is not filtered");
}
#[test]
fn wire_response_parses_f4_shape() {
let raw = r#"{
"data": [
{
"id": "kimi-for-coding",
"context_length": 262144,
"supports_reasoning": true,
"supports_image_in": true,
"supports_video_in": false,
"display_name": "k2.6-code-preview"
},
{ "id": "kimi-k2-turbo-preview" }
]
}"#;
let resp: WireModelsResponse = serde_json::from_str(raw).expect("F4 shape must parse");
assert_eq!(resp.data.len(), 2);
let first = &resp.data[0];
assert_eq!(first.id, "kimi-for-coding");
assert_eq!(first.context_length, 262_144);
assert_eq!(first.display_name.as_deref(), Some("k2.6-code-preview"));
assert_eq!(
first.capabilities(),
vec![ModelCapability::Thinking, ModelCapability::ImageIn]
);
// Missing optional fields default off/0.
let second = &resp.data[1];
assert_eq!(second.context_length, 0);
assert!(!second.supports_reasoning);
}
#[test]
fn bundled_fallback_is_kimi_catalog() {
assert_eq!(default_model(), "kimi-for-coding");
// Aux models fall back to the default (no dedicated entries).
assert_eq!(default_image_description_model(), "kimi-for-coding");
assert_eq!(default_session_summary_model(), "kimi-for-coding");
// No kigi remnants in the embedded fallback.
assert!(!DEFAULT_MODELS_JSON.contains("kigi"));
}
}
Reference in New Issue View Git Blame Copy Permalink