Models API (Anthropic Compatible)¶

Discover and list available Amazon Bedrock text models through an Anthropic-compatible interface. Served under /anthropic by default; the examples below use $BASE, which includes that prefix.

At a glance¶

Text model catalog — Browse the available Bedrock text models across regions. See the Models page for a browsable table.
Always up to date — Dynamic model discovery automatically shows new models as they become available in Bedrock.
Multi-region aggregation — Combines models from all configured AWS regions in one list. See which models are available in each region.
Foundation models — Includes Claude, Nova, Llama, and other Bedrock foundation text models.
Differs from the Anthropic API: model IDs are Bedrock identifiers, the catalogue is not limited to Claude, and limit defaults to 1000 instead of 20 — see Limits and behaviour to know.
Only text-in, text-out models are listed. For embedding, image, audio or video models, and for filtering by route, region or MCP tool, use GET /search_models.

Base URL and route prefix

By default, all Anthropic-compatible routes are prefixed with /anthropic. This means the Models API is available at /anthropic/v1/models instead of /v1/models. You can customize this prefix using the ANTHROPIC_ROUTES_PREFIX configuration variable documented in HTTP Server and MCP.

The curl examples on this page use a $BASE variable that must include this prefix — set it to your scheme and host followed by ANTHROPIC_ROUTES_PREFIX:

export BASE="https://your-host/anthropic"  # <scheme>://<host> + ANTHROPIC_ROUTES_PREFIX

Note: ANTHROPIC_ROUTES_PREFIX must always be a non-empty path and is validated at startup to differ from OPENAI_ROUTES_PREFIX (the server refuses to start otherwise). This Anthropic-compatible Models API is therefore always served on its own path, distinct from the OpenAI-compatible Models API.

curl "$BASE/v1/models" -H "x-api-key: $ANTHROPIC_API_KEY"

Endpoints¶

Endpoint	Method	What It Does	Powered By	MCP Tool
`/v1/models`	`GET`	List all available text models	Amazon Bedrock	`anthropic_model_list`
`/v1/models/{model_id}`	`GET`	Get details for a specific model	Amazon Bedrock	`anthropic_model_get`

Working with the catalogue¶

Features:

Multi-region aggregation: Combines models from all configured Bedrock regions
Cursor-based pagination: Use limit, after_id, and before_id query parameters
Text models only: Returns only models with text input and text output modalities (Claude, Nova, Llama, etc.)

Response fields¶

GET /v1/models returns a paginated envelope; GET /v1/models/{model_id} returns one model object with the same four fields.

Field	Type	Description
`data[]`	array	The page of models, sorted by model ID
`id`	string	Bedrock model identifier, the value to send as `model`
`type`	string	Always `model`
`display_name`	string	Human-readable model name
`created_at`	string	RFC 3339 release date — see the note below
`has_more`	boolean	Whether another page follows the cursor
`first_id`	string \| null	ID of the first model on this page, to use as `before_id`
`last_id`	string \| null	ID of the last model on this page, to use as `after_id`

{
  "data": [
    {
      "id": "amazon.nova-micro-v1:0",
      "type": "model",
      "display_name": "Amazon Nova Micro",
      "created_at": "2025-01-01T00:00:00Z"
    }
  ],
  "has_more": false,
  "first_id": "amazon.nova-micro-v1:0",
  "last_id": "amazon.nova-micro-v1:0"
}

Created Date (created_at)

The created_at field is an RFC 3339 datetime string representing the time at which the model was released. This value is sourced from the Bedrock model lifecycle metadata (startOfLifeTime). If the release date is not available from Bedrock, it defaults to the Unix epoch ("1970-01-01T00:00:00Z").

Limits and behaviour to know¶

What's Different from Anthropic?¶

Model IDs: Uses Bedrock model identifiers (e.g., anthropic.claude-haiku-4-5-20251001-v1:0) instead of Anthropic model names
Extended catalog: Includes all Bedrock text models (Claude, Nova, Llama, etc.), not just Anthropic models
Default page size: limit defaults to 1000 (the Anthropic API defaults to 20)

Only text-in, text-out models appear. A model is listed when its Bedrock metadata declares both TEXT input and TEXT output, so embedding, image, speech and video models are absent from this catalogue even though the deployment serves them. GET /search_models lists every model with its modalities and routes.

The listing carries no capability metadata. id, type, display_name and created_at are the whole object: there is no context window, no pricing, no region and no tool support. Those come from GET /search_models and GET /model_pricing.

Request headers¶

Header	Purpose	Notes
`x-api-key`	Gateway API key	Required, like every other route
`anthropic-version`	Anthropic API version pin	Accepted and ignored; sent by the official SDKs

Try it¶

List all available models:

curl -X GET "$BASE/v1/models" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01"

List models with pagination:

curl -X GET "$BASE/v1/models?limit=10" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01"

The limit query parameter accepts values from 1 to 1000 and defaults to 1000.

Get details for a specific model:

curl -X GET "$BASE/v1/models/amazon.nova-micro-v1:0" \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H "anthropic-version: 2023-06-01"

Next steps¶

Next: Messages API · Search Models API · Model Pricing API · Models catalogue