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

# Models

> GET /v1/models - list verified deployed models in your workspace.

```http theme={"dark"}
GET https://api.runinfra.ai/v1/models
```

Returns verified deployed models across all **active\_verified** deployments in your workspace. The shape matches OpenAI's `client.models.list()` response for supported model-list clients.

## Request

<RequestExample>
  ```python Python theme={"dark"}
  from openai import OpenAI
  client = OpenAI(base_url="https://api.runinfra.ai/v1", api_key="YOUR_RUNINFRA_API_KEY")

  for model in client.models.list().data:
      print(model.id, model.owned_by)
  ```

  ```javascript JavaScript theme={"dark"}
  const models = await client.models.list();
  models.data.forEach((m) => console.log(m.id, m.owned_by));
  ```

  ```bash cURL theme={"dark"}
  curl https://api.runinfra.ai/v1/models \
    -H "Authorization: Bearer YOUR_RUNINFRA_API_KEY"
  ```
</RequestExample>

## Response

```json theme={"dark"}
{
  "object": "list",
  "data": [
    {
      "id": "llama-3.3-70b",
      "object": "model",
      "created": 1714502400,
      "owned_by": "runinfra"
    },
    {
      "id": "bge-m3",
      "object": "model",
      "created": 1714502400,
      "owned_by": "runinfra"
    }
  ]
}
```

## Semantics

* **De-duplicated** - the same `modelId` deployed in two different verified pipelines appears only once.
* **Live** - reflects verified deployment state; stopped, failed, health-only, or unverified deployments disappear from the list within seconds.
* **No credits charged** - model discovery is free; rate-limit budget is not consumed by this endpoint so SDK boot flows that call it eagerly don't inflate your quota usage.
* **Pipeline-scoped keys** see only their pipeline's verified models; workspace-scoped keys see verified deployed models in the workspace.

## Notes

<Note>
  The `created` timestamp is cosmetic - it reflects the response time, not the deployment's actual create time. SDK clients generally only rely on `id` and `object` from this payload.
</Note>

<Warning>
  Models your pipeline uses but hasn't deployed (e.g., listed in the Model tab but with status `stopped`) do NOT appear here. This endpoint reflects runtime reachability, not pipeline configuration.
</Warning>