InferenceReference

Models

A method that returns the list of models on your plan along with the context length and capabilities of each.

The Models method returns the list of models available on your plan and details about each: type, context length, maximum reply length and capabilities. It is useful for finding the exact model name for the model field and checking in advance what the model can do. Models that are not included in your plan do not appear in the response.

Request

curl https://api.yenisei.ru/v1/models \
  -H "Authorization: Bearer $YENISEI_KEY"

Response

The example is shortened: each model has other fields too, and they are listed below.

{
  "object": "list",
  "data": [
    {
      "id": "yenisei-100",
      "object": "model",
      "publisher": "Yenisei",
      "type": "chat",
      "context_length": 1000000,
      "capabilities": { "reasoning": false, "tools": true }
    },
    {
      "id": "bge-m3",
      "object": "model",
      "publisher": "BAAI",
      "type": "embedding",
      "context_length": 8192,
      "capabilities": {}
    }
  ]
}

Model fields

• idstring

  Model name. This is what you put in the model field of a request.

• typestring

  Model type: chat for a conversation model, embedding for an embedding model.

• publisherstring

  Who released the model. If the catalog does not specify it, the field is absent from the response.

• context_lengthinteger

  How many tokens the model accepts per request, counting the prompt and the reply together.

• max_output_tokensinteger

  Maximum reply length in tokens. A max_tokens value larger than this number is cut down to it.

• capabilitiesobject

  Model capabilities from the catalog, for example {"reasoning": true}.

The list depends on your plan: a model that is not included in your plan will not appear in the response. Model descriptions and prices are collected in the catalog.