Models

List models

curl https://api.avenro.tech/v1/models

{
  "object": "list",
  "data": [
    {
      "id": "gpt-6-astra",
      "object": "model",
      "created": 1791331200,
      "owned_by": "OpenAI",
      "context_length": 1050000,
      "max_output_tokens": 128000,
      "pricing": {
        "unit": "USD per million tokens",
        "input": "6",
        "cached_input": null,
        "output": "30",
        "long_context": {
          "above_input_tokens": 272000,
          "input": "12",
          "cached_input": null,
          "output": "45"
        },
        "valid_until": null
      }
    }
  ]
}

• No API key is needed.
• Only models that can serve requests at this moment are listed. The shape matches OpenAI's, so client.models.list() works in the OpenAI SDKs.
• context_length, max_output_tokens and pricing are Avenro additions. Prices are decimal strings in US dollars per million tokens; cached_input is null when the model has no separate cached price.
• long_context is null unless the provider bills long prompts at higher prices: a request with more than above_input_tokens prompt tokens is billed at those prices for all of its tokens, as the provider bills it.
• valid_until is null unless the provider has announced a price change. It is the Unix time at which the listed prices stop holding; from then the model is unavailable until its new prices are posted.

Retrieve a model

curl https://api.avenro.tech/v1/models/gpt-6-astra

Returns one entry in the same shape, or 404 model_not_found. The models page shows the same catalog with the reference prices savings are measured against.