List Models

GET /admin/v1/models

List Models

List all configured model resources.

Parameters

Responses

• 200: OK
• [].id (string, required): Resource ID, as assigned by the active resource source.
• [].revision (integer, required): Monotonic resource revision: the etcd mod_revision of the entry, or the load generation in file mode.
• [].value (object, required): Stored model configuration.
• allOf variant 1
• [].value.allowed_cidrs (array, optional): Client IP allowlist in CIDR notation. Empty or absent allows all clients.
• [].value.auto_prompt_caching (object, optional): Automatic prompt-cache breakpoint injection for direct Anthropic models. Omit to leave injection off.
• allOf variant 1
• [].value.auto_prompt_caching.enabled (boolean, required): Whether automatic prompt-cache injection is active for this model.
• [].value.auto_prompt_caching.ttl (object, optional): Cache lifetime for injected breakpoints: 5m (default when omitted) or 1h. A 1h cache write costs 2x the base input rate versus 1.25x for 5m, so it pays off only when the cached prefix is reused across a longer session.
• allOf variant 1
• [].value.background_model_check (object, optional): Direct-model-only background health-check configuration.
• allOf variant 1
• [].value.background_model_check.enabled (boolean, required): Whether background health checks are enabled for this model.
• [].value.background_model_check.ignore_statuses (array, optional): Upstream status codes to ignore when evaluating background check failures.
• [].value.background_model_check.interval_seconds (integer, required): Seconds between background health checks. Minimum: 5.
• [].value.background_model_check.max_tokens (integer, required): Maximum completion tokens requested during each background health check.
• [].value.background_model_check.prompt (string, required): Prompt sent to the model during each background health check.
• [].value.background_model_check.stale_after_seconds (integer, required): Seconds after which the last completed background check is considered stale.
• [].value.background_model_check.timeout_seconds (integer, required): Request timeout in seconds for each background health check. Minimum: 1.
• [].value.cooldown (object, optional): Direct-model-only request-path cooldown configuration. Cooldown is opt-in: omit this field, or leave enabled unset, and the model is never taken out of rotation by request-path failures.
• allOf variant 1
• [].value.cooldown.default_seconds (integer, optional): Cooldown TTL in seconds when the upstream did not supply a Retry-After header or honor_retry_after is false.
• [].value.cooldown.enabled (boolean, optional): Whether cooldown is active for this model. Cooldown is off unless this is set to true, so a model that omits it stays in rotation regardless of upstream failures.
• [].value.cooldown.honor_retry_after (boolean, optional): Whether to use the upstream's Retry-After header as the cooldown TTL when it contains seconds.
• [].value.cooldown.max_seconds (integer, optional): Upper bound on cooldown TTL when Retry-After is used.
• [].value.cooldown.trigger_on_timeout (boolean, optional): Whether request-path timeouts trigger cooldown.
• [].value.cooldown.trigger_on_transport (boolean, optional): Whether transport, decode, or stream-abort errors trigger cooldown.
• [].value.cooldown.trigger_statuses (array, optional): Status codes that trigger cooldown, covering authentication failures, rate limits, and transient server errors. Caller-side validation errors such as 400, 403, and 422 are excluded.
• [].value.cost (object, optional): Per-token cost for budget tracking. Omit it when cost tracking is not needed.
• allOf variant 1
• [].value.cost.input_per_1k (number, required): Prompt token cost in USD per 1,000 tokens.
• [].value.cost.output_per_1k (number, required): Completion token cost in USD per 1,000 tokens.
• [].value.display_name (string, required): Operator-facing unique label. Surfaces on /v1/models, req.model on chat completions, ApiKey.allowed_models, and the dashboard model list. Resource::name() returns this.
• [].value.effort_mapping (object, optional): Direct-model-only mapping from a client-requested reasoning effort to the value sent upstream. The gateway applies one exact lookup after resolving the final target; unlisted values pass through unchanged.
• [].value.embedding (object, optional): Embedding-modality metadata. Present on direct Models that serve an OpenAI-compatible /v1/embeddings endpoint (and can be referenced by a semantic router's embedding_model).
• allOf variant 1
• [].value.embedding.dimensions (integer, required): Output vector dimensionality. Used to validate vectors, key the example-vector cache, and (for endpoints that support it) request a reduced output size.
• [].value.embedding.normalize (boolean, optional): Whether the endpoint already returns L2-normalized vectors. When false, the gateway normalizes before computing cosine similarity. Defaults to true.
• [].value.ensemble (object, optional): Ensemble configuration for panel calls and judge synthesis.
• allOf variant 1
• [].value.ensemble.judge (object, required): Direct model that combines successful panel responses.
• allOf variant 1
• [].value.ensemble.judge.model (string, required): Model alias for the direct model that synthesizes panel responses.
• [].value.ensemble.judge.synthesis_prompt (string, optional): Override for the built-in synthesis prompt template.
• [].value.ensemble.min_responses (integer, optional): Minimum successful panel responses required before judge synthesis. When omitted, the gateway requires the smaller of 2 and the panel size.
• [].value.ensemble.panel (array, required): Direct models called concurrently for each ensemble request.
• [].value.ensemble.panel[].model (string, required): Model alias for a direct model that receives one panel request.
• [].value.ensemble.panel[].seed (integer, optional): Sampling seed for this panel member.
• [].value.ensemble.panel[].temperature (number, optional): Sampling temperature for this panel member. Omit it to keep the request's temperature.
• [].value.ensemble.panel[].weight (integer, optional): Reserved for a future voting/quorum strategy. AISIX currently ignores this field.
• [].value.ensemble.timeout_ms (integer, optional): Per-call upstream deadline applied to each panel member and the judge. Set 0 or omit it to disable the ensemble-level deadline.
• [].value.model_name (string, optional): Upstream model identifier sent in provider requests. Routing and ensemble models leave this field unset.
• [].value.provider (string, optional): Upstream vendor identity used for dispatch, compatibility checks, telemetry, and access logs. Routing and ensemble models leave this field unset.
• [].value.provider_key_id (string, optional): Provider key resource ID used to authenticate upstream requests. Routing and ensemble models leave this field unset.
• [].value.rate_limit (object, optional): Request, token, and concurrency limits for this model.
• allOf variant 1
• [].value.rate_limit.concurrency (integer,null, optional): Max concurrent in-flight requests.
• [].value.rate_limit.rpd (integer,null, optional): Requests per 86,400-second window.
• [].value.rate_limit.rph (integer,null, optional): Requests per 3,600-second window. There is no per-hour token limit field.
• [].value.rate_limit.rpm (integer,null, optional): Requests per 60-second window.
• [].value.rate_limit.rps (integer,null, optional): Requests per 1-second window. There is no per-second token limit field.
• [].value.rate_limit.tpd (integer,null, optional): Tokens per 86,400-second window.
• [].value.rate_limit.tpm (integer,null, optional): Tokens per 60-second window.
• [].value.retries (integer, optional): Retry attempts against this model after a retryable upstream failure, before the request gives up (or, inside a model group, fails over to the next target). Absent falls back to the group's routing.retries, then to the deployment-wide upstream.retries default.
• [].value.routing (object, optional): Virtual routing configuration. When set, the gateway selects a target from routing.targets and uses that target model's provider, model_name, and provider_key_id fields for upstream dispatch.
• allOf variant 1
• [].value.routing.fallback_on_statuses (array,null, optional): Additional upstream HTTP status codes that participate in retries and failover. By default a non-429 4xx response is treated as a caller error and returned as-is; providers that use 4xx codes for transient conditions (model overload, queue full, quota exhaustion) can be listed here, for example [408, 409]. 5xx codes are already retryable, so listing them changes nothing. Authentication (401/403) and validation (400) codes should only be listed when the provider is known to use them for transient failures.
• [].value.routing.hash_on (array,null, optional): Where the consistent_hash hash key comes from: an ordered chain of sources, the first non-empty value winning. Defaults to the x-aisix-routing-key request header, falling back to the caller's API key id. Only valid with strategy: consistent_hash.
• [].value.routing.hash_on[].name (string, optional): The header or cookie name to read. Required for header and cookie sources; not accepted for api_key or client_ip.
• [].value.routing.hash_on[].type (object, required): Which request attribute supplies the hash key.
• allOf variant 1
• oneOf variant 1: Header
• oneOf variant 2: Cookie
• oneOf variant 3: API key
• oneOf variant 4: Client IP
• [].value.routing.max_fallbacks (integer,null, optional): Max number of later targets to attempt after the initial target fails permanently. When omitted, all later targets may be attempted.
• [].value.routing.retries (integer,null, optional): Retry attempts on the current target before failing over, applied to every target that does not set its own retries. Absent falls back to the deployment-wide upstream.retries default.
• [].value.routing.retry_on_429 (boolean,null, optional): Whether upstream 429 participates in retries and failover.
• [].value.routing.strategy (object, optional): Strategy used to select a target for each request.
• allOf variant 1
• oneOf variant 1: Round robin
• oneOf variant 2: Consistent hash
• oneOf variant 3: Failover
• oneOf variant 4: Least cost
• oneOf variant 5: Least latency
• oneOf variant 6: Least busy
• [].value.routing.targets (array, required): Ordered set of direct models available to this routing model.
• [].value.routing.targets[].model (string, required): Model alias for a direct model that can receive routed traffic.
• [].value.routing.targets[].priority (integer, optional): Priority tier, default 0; a higher value is preferred (the APISIX node-priority convention — give backup targets -1). The strategy orders targets within each tier; a lower tier is only tried when every higher-tier target failed or is unavailable.
• [].value.routing.targets[].tags (array, optional): Tags for tag/metadata-conditional routing. When a request carries routing tags, only targets whose tags intersect the request's are eligible; a target tagged "default" is the fallback used when nothing matches and for untagged requests. Absent/empty means the target opts out of tag filtering (eligible only via the default fallback once any sibling target is tagged). The configured strategy then orders whatever set survives.
• [].value.routing.targets[].weight (integer, optional): Target weight, default 1. Used by round_robin (rotation share), consistent_hash (share of the hash ring), and least_busy (in-flight divided by weight). failover, least_cost, and least_latency accept the field but do not use it.
• [].value.routing.when_all_unavailable (object, optional): Policy to apply when every target is unavailable because of runtime health or cooldown state.
• allOf variant 1
• oneOf variant 1: Fail
• oneOf variant 2: Try anyway
• [].value.semantic (object, optional): Semantic-routing configuration. When set, the gateway embeds the request and dispatches to the route whose examples it matches best, using that route's target Model for upstream dispatch.
• allOf variant 1
• [].value.semantic.default (string, required): Direct model alias used when no route clears its threshold.
• [].value.semantic.embedding_model (string, required): Alias of an embedding-modality Model used to embed the request and (at apply time) the route examples.
• [].value.semantic.embedding_timeout_ms (integer, optional): Per-call deadline for the embedding request in milliseconds. 0 or absent disables the embedding-specific deadline.
• [].value.semantic.match (object, required): Shared matching parameters (metric, aggregation, default threshold).
• allOf variant 1
• [].value.semantic.match.aggregation (object, optional): Per-example score aggregation. v1: max.
• allOf variant 1
• oneOf variant 1: Max
• [].value.semantic.match.distance_metric (object, optional): Similarity metric. v1: cosine.
• allOf variant 1
• oneOf variant 1: Cosine
• [].value.semantic.match.threshold (number, required): Default similarity threshold for routes that do not set their own threshold. Higher is stricter.
• [].value.semantic.on_embedding_failure (object, optional): Behavior when the embedding call fails or times out. Defaults to routing to default.
• allOf variant 1
• anyOf variant 1: Default or fail policy
• allOf variant 1
• oneOf variant 1: Default
• oneOf variant 2: Fail
• anyOf variant 2: Target model
• [].value.semantic.on_embedding_failure.target (string, required): Direct-model alias to route to when embedding fails.
• [].value.semantic.routes (array, required): Routes evaluated for each request. At least one is required.
• [].value.semantic.routes[].description (string, optional): Human-facing description. Documentation only — v1 matches on examples, not on this field.
• [].value.semantic.routes[].examples (array, required): Example utterances that define this route. AISIX embeds each example when applying the configuration and caches the vector. A request is matched against these examples. At least one example is required.
• [].value.semantic.routes[].name (string, required): Operator-facing route label. Surfaced in the x-aisix-route response header and access logs (e.g. prod-chat -> route:legal).
• [].value.semantic.routes[].target (string, required): Direct model alias that receives traffic matching this route.
• [].value.semantic.routes[].threshold (number, optional): Per-route similarity threshold. A request matches this route only when its aggregated score is >= this value. When omitted, the router-level threshold applies.
• [].value.stream_timeout (integer, optional): Maximum gap in milliseconds between upstream streaming chunks. 0 or absent falls back to the group's stream_timeout, then to the model's (or group's) timeout, then to the deployment-wide upstream.stream_timeout_ms / timeout_ms defaults.
• [].value.timeout (integer, optional): End-to-end timeout in milliseconds for non-streaming upstream calls. Absent falls back to the group's timeout, then to the deployment-wide upstream.timeout_ms default. 0 disables the non-streaming timeout for this model.
• oneOf variant 1: Routing model
• oneOf variant 2: Direct model
• oneOf variant 3: Ensemble model
• oneOf variant 4: Semantic router
• 401: Missing or invalid admin key
• error_msg (string, required): Human-readable error message.
• 500: Configuration store operation failed Response fields (application/json): identical to GET /admin/v1/models, response 401 above.

cURL

curl -X GET 'http://127.0.0.1:3001/admin/v1/models'