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

# Models

> Discover available model IDs and select one for your request.

Call `GET /v1/models` with a customer key to obtain the current catalog:

```bash theme={null}
curl --fail-with-body "$LIMITLESS_BASE_URL/models" \
  -H "Authorization: Bearer $LIMITLESS_API_KEY"
```

The response uses the OpenAI model-list structure:

```json theme={null}
{
  "object": "list",
  "data": [
    {
      "id": "abliterated-model-large-v2",
      "object": "model",
      "created": 0,
      "owned_by": "provider"
    }
  ]
}
```

This is an illustrative response. Model IDs and metadata come from the configured backend; use your actual response as the source of available IDs. A catalog entry does not guarantee that every optional API feature is supported or that billing is enabled for that model.

## Current console choices

| Display name | API model ID |
| - | - |
| Abliterated | `abliterated-model` |
| GLM 5.3 | `abliterated-model-large-v2` |
| GLM 5.2 | `abliterated-model-large` |

The console choices reflect this release's configured model list. Check the live catalog before depending on any model ID.

## Retrieve one model

```bash theme={null}
curl --fail-with-body "$LIMITLESS_BASE_URL/models/abliterated-model-large-v2" \
  -H "Authorization: Bearer $LIMITLESS_API_KEY"
```

An unavailable ID returns `404`. Model discovery does not generate tokens or create an inference usage event.

## Choose settings deliberately

Use modest output limits for initial tests. Tool calling, structured output, reasoning parameters, and multimodal inputs are model-specific. There is no universal context-window or feature guarantee across the catalog.

See [compatibility](/guides/compatibility) and [billing](/billing) before enabling a new model in production.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.