> ## Documentation Index
> Fetch the complete documentation index at: https://docs.geminix.cc/llms.txt
> Use this file to discover all available pages before exploring further.

# Models

> List available models and set the correct model field on requests.

Every gateway request must use a `model` (or path model id) that your account can access. Availability depends on channels and groups configured by admins.

## List models

```bash theme={null}
curl https://api.geminix.cc/v1/models \
  -H "Authorization: Bearer $OMNIMUX_API_KEY"
```

The response follows the OpenAI models list shape. You can also browse models in the console under **Models / Channels** (menu labels vary by deployment).

Gemini-style list:

```bash theme={null}
curl https://api.geminix.cc/v1beta/models \
  -H "Authorization: Bearer $OMNIMUX_API_KEY"
```

## Use a model in a request

Chat example:

```bash theme={null}
curl https://api.geminix.cc/v1/chat/completions \
  -H "Authorization: Bearer $OMNIMUX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "ping"}]
  }'
```

Replace `gpt-4o-mini` with a real `id` from your list.

## Protocol notes

| Protocol         | Example path                                  | Where model is set |
| ---------------- | --------------------------------------------- | ------------------ |
| OpenAI Chat      | `POST /v1/chat/completions`                   | JSON `model`       |
| OpenAI Responses | `POST /v1/responses`                          | JSON `model`       |
| Claude Messages  | `POST /v1/messages`                           | JSON `model`       |
| Gemini           | `POST /v1beta/models/{model}:generateContent` | Path `{model}`     |

Full schemas live in the [API reference](/en/api-reference/overview).

## When a model fails

1. Token model allowlist too narrow
2. Account group does not include the model
3. Upstream channel disabled
4. Exact id mismatch (case-sensitive)