# GPT API and pricing (https://wanapi.app/docs/gpt-api)



Use GPT models with the same API key and prepaid balance as the other models on Wan API. Each model has its own parameter reference and live pricing table.

## Connect [#connect]

Create a key in [API key settings](/settings/apikeys) and keep it on your server. Set the OpenAI SDK base URL to `https://api.wanapi.app/v1`.

| Protocol                | Endpoint                    | Authentication                       |
| ----------------------- | --------------------------- | ------------------------------------ |
| OpenAI Responses        | `POST /v1/responses`        | `Authorization: Bearer YOUR_API_KEY` |
| OpenAI Chat Completions | `POST /v1/chat/completions` | `Authorization: Bearer YOUR_API_KEY` |

Send JSON with `Content-Type: application/json`. Responses arrive directly or as SSE; there is no video-style task to poll.

## Models [#models]

| Model                                | Request model ID | Reasoning                              |
| ------------------------------------ | ---------------- | -------------------------------------- |
| [GPT Luna 6](/docs/gpt-6-luna)       | `gpt-6-luna`     | Supports `none`; default `medium`.     |
| [GPT Terra 6.1](/docs/gpt-6-1-terra) | `gpt-6.1-terra`  | Supports `none`; default `medium`.     |
| [GPT Sol 6.1](/docs/gpt-6-1-sol)     | `gpt-6.1-sol`    | Always reasons; default `medium`.      |
| [GPT Astra 6](/docs/gpt-6-astra)     | `gpt-6-astra`    | Always reasons; set effort explicitly. |

All four accept text and images and produce text. Use the exact request ID, including decimal points in 6.1. Authenticated `GET /v1/models` lists models currently available to your key. A documentation page alone does not establish channel availability.

## Groups and pricing [#groups-and-pricing]

One key works across models and groups. Requests route automatically to eligible groups and are billed at the rates of the group actually selected. The model's price table includes each group's multiplier; **Default** marks the first eligible group in the configured routing order.

Requests without a group header use `gpt`. To choose a group for Responses or Chat Completions, send `X-Model-Group: gpt` or `X-Model-Group: gpt-0.7`.

`gpt-0.7` is available for Luna, Sol and Astra only; Terra supports `gpt` only. Explicit selections never fall back to another group. Keep the model ID unchanged; group suffixes are not supported.

Missing or unrecognized rates are shown as unavailable, never as free. Review [LLM logs](/usage) after a call for the actual usage and charge.
