API reference

Models

GET /v1/models and GET /v1/models/{id}

List models#

endpoint

GET https://api.stardustllm.xyz/v1/models

Returns every model you can call. No API key is needed. The catalog changes as models are added, so read it instead of hard-coding ids.

response (shortened)

{
  "object": "list",
  "data": [
    {
      "id": "glm-5.3-flash-free",
      "object": "model",
      "name": "GLM-5.3-Flash Free",
      "vendor": "Z.ai",
      "context_length": 200000,
      "max_output": 32000,
      "input_modalities": ["text", "image", "video"],
      "pricing": { "input": 0, "output": 0 },
      "free": true,
      "tools": false,
      "reasoning": true,
      "open_weight": false,
      "original": { "id": "z-ai/glm-5.3-flash", "context_length": 1048576 }
    }
  ]
}

Fields#

Field	Description
`id`	The value to send as `model`.
`context_length`	Maximum tokens of input + output.
`max_output`	Maximum tokens in one answer.
`input_modalities`	What the model accepts, e.g. `text`, `image`.
`pricing`	USD per 1M tokens for `input` and `output`. Discounted models also show `list_input`, `list_output` and `discount_percent`.
`free`	`true` for free models (need 5,000 ENERGY held, spend nothing).
`tools`, `reasoning`	Whether the model supports tool calls and reasoning.
`open_weight`	Whether the model's weights are public.
`original`	For free variants: the original model id and its full context length.

Model profile#

endpoint

GET https://api.stardustllm.xyz/v1/models/{id}

Adds the description, release date, the model a free variant is based_on, free_min_energy, cache pricing and the providers serving it. The same data is shown on each model page of the catalog.