Developer reference

API documentation

Use one API Key and Base URL to access the models published by this gateway.

Connect

Base URL: /v1
Authentication: Authorization: Bearer $TOKENSSS_API_KEY

Model

No public models are visible yet

Browse models

First request

The example follows the selected public model automatically.

curl '/v1/chat/completions' \
  -H "Authorization: Bearer ${TOKENSSS_API_KEY}" \
  -H 'Content-Type: application/json' \
  --data-raw '{
  "model": "your-model-id",
  "messages": [
    {
      "role": "user",
      "content": "Hello"
    }
  ]
}'

Supported endpoints

POST /v1/responses

Create Responses API outputs with streaming and tools support.

POST /v1/chat/completions

Create a model response from a chat message sequence.

POST /v1/embeddings

Create embeddings for text inputs.

POST /v1/images/generations

Generate images from text prompts.

POST /v1/audio/transcriptions

Transcribe audio with an OpenAI-compatible request.

GET /v1/models

List models published by this gateway.

POST /v1/messages

Send Anthropic Messages-compatible requests.

POST /v1beta/models/:model:generateContent

Send Gemini-native generateContent requests.

Streaming

Set stream to true where the endpoint supports it. Consume Server-Sent Events until the terminal event or marker arrives.

Errors and retries

Errors use the gateway error envelope and preserve actionable upstream context when available. Retry only transient failures.