Models

Catalog

Live models can be called now. Coming-soon entries are listed so you can plan; the API returns 404 for them until they are served. Prices are Thai baht per one million tokens. Image, video and audio models will get their own OpenAI-style endpoints and per-output pricing at launch; they are never routed through chat completions.

Model ids

Ids are namespaced thairouter/<model>. Send the full id in model. Responses echo the same id, never the upstream weight name. Ids are stable; when a newer revision replaces a model we add a new id rather than changing what an existing one points at.

Listing models

Returns only live, routable models in the OpenAI list shape, with ThaiRouter extras per entry: limits, pricing, the reasoning levels the model distinguishes (null for non-reasoning models) and the parameters it accepts.

{
  "object": "list",
  "data": [
    {
      "id": "thairouter/glm-5.3-flash",
      "object": "model",
      "created": 0,
      "owned_by": "thairouter",
      "context_length": 393216,
      "max_output_tokens": 131072,
      "pricing": { "promptPerMillion": 5, "completionPerMillion": 15 },
      "reasoning": { "efforts": ["low", "high", "max"], "default": "max", "optional": false },
      "supported_parameters": ["max_tokens", "temperature", "top_p", "stop", "seed",
        "frequency_penalty", "presence_penalty", "response_format", "stream",
        "reasoning", "reasoning_effort"]
    }
  ]
}

Unauthenticated catalog data, including coming-soon models and live performance stats, is available at GET /public/models. See the API reference.

Limits