Model list

Naming

A model ID is simply the model's own name, with no vendor prefix, for example:

claude-opus-4-7
gpt-5.6-sol
gemini-2.5-pro
deepseek-v4-flash
kimi-k3

One model can be served by several upstream providers; routing picks one, and that choice never appears in the ID. IDs are case-sensitive — use them exactly as the API returns them.

Fetching the list

curl https://api.apione.ai/v1/models \
  -H "Authorization: Bearer $APIONE_API_KEY"

{
  "object": "list",
  "data": [
    { "id": "claude-opus-4-7", "object": "model", "created": 0, "owned_by": "maxx" },
    { "id": "gpt-5.6-sol", "object": "model", "created": 0, "owned_by": "maxx" }
  ]
}

The list covers models this key may call; if the key is limited to selected models, only those appear. For one model's supported parameters and billing mode, call GET /v1/models/{model}/params — see Reference › GET /models.

Browsing on the web

The model catalog lists every available model, each with pricing, context length, supported inputs and capabilities. It is easier to compare there than in JSON.

Three rules of thumb

• Tier by task: high-volume short work — classification, extraction, rewriting — goes to small models (claude-haiku-4-5-20251001, gemini-3.5-flash); save flagships for hard reasoning and long documents.
• Look at context, not size: if a whole manual has to fit, check the context length first — overflowing returns context_length_exceeded.
• Compare on a small batch: run the same prompts through two or three models and compare quality and usage before committing.

Additions and retirements

New models work the moment they launch — no configuration on your side. Retirements are announced in advance; calling a retired model returns 400 model_not_supported.