Create a message (Anthropic Messages format)

POST

/messages

Accepts the Anthropic Messages request format and routes only to models that list messages
in their catalog endpoints. The data plane requires a JSON object and a non-empty model;
the provider validates the rest of the body. extra.fallbacks is consumed and extra is
removed before forwarding. The response id is rewritten to the gateway request ID and
model to the public catalog model ID.

Authentication: Authorization: Bearer <key> (default). x-api-key is accepted only when no
Authorization header is sent; an empty or invalid Bearer value returns 401 and does not
fall back to x-api-key.

With stream: true the response is the Anthropic SSE event stream (message_start,
content_block_start, content_block_delta, content_block_stop, message_delta,
message_stop); tool input may arrive in input_json_delta pieces. This endpoint does not
return the OpenGateway debug object. Errors use the Anthropic error envelope.

Request

Header Params

Body Params application/jsonRequired

Example

{
    "model": "deepseek/deepseek-v4.1-flash-ultrafast",
    "max_tokens": 256,
    "messages": [
        {
            "role": "user",
            "content": "Hello"
        }
    ]
}

Request Code Samples

Shell

JavaScript

Java

Swift

PHP

Python

HTTP

Objective-C

Ruby

OCaml

Dart

curl --location '/messages' \
--header 'anthropic-version;' \
--header 'anthropic-beta;' \
--header 'x-opengateway-user-id;' \
--header 'x-opengateway-session-id;' \
--header 'x-sionic-task-id;' \
--header 'Content-Type: application/json' \
--data '{
    "model": "deepseek/deepseek-v4.1-flash-ultrafast",
    "max_tokens": 256,
    "messages": [
        {
            "role": "user",
            "content": "Hello"
        }
    ]
}'

Responses

🟢200OK

application/json

An Anthropic Message object, or the Anthropic SSE event stream when stream is true.

Body

Example

{
    "id": "msg_example",
    "type": "message",
    "role": "assistant",
    "model": "deepseek/deepseek-v4.1-flash-ultrafast",
    "content": [
        {
            "type": "text",
            "text": "Hello!"
        }
    ],
    "stop_reason": "end_turn",
    "stop_sequence": null,
    "usage": {
        "input_tokens": 12,
        "output_tokens": 4
    }
}

🟠400Bad Request

🟠401Unauthorized

🟠402

🟠403Forbidden

🟠404Record Not Found

🟠429

🔴500Server Error

🔴503Service Unavailable

🔴504Gateway Timeout

🔴529

Modified at 2026-10-27 15:03:30

Create a model response

Create embeddings