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

# Models

> Discover available models and ask Gumloop Chew to choose one for a task.

```bash theme={"dark"}
gumloop models list --json
gumloop models route "Summarize this report" --json
gumloop models route "Review this code" --model MODEL_ID --model OTHER_MODEL_ID --json
```

`list` returns model metadata and availability for your account. Use returned model IDs in requests. `route` asks Gumloop Chew to select a model for the task; it does not run the task. Routing can consume credits.

With the corresponding backend discovery update, the catalog also includes registered image models and callable decision-model IDs. Availability and policy flags still apply; discovery does not grant access to a model.

Pass `--input-stdin -` to read a task from stdin. `--system-prompt` or `--system-prompt-file` supplies the agent's standing instructions for routing. Run `gumloop models route --help` for the remaining agent-context options.

Use [`gumloop chat completions create`](https://docs.gumloop.com/cli/chat) for an independent model call, or [`gumloop sessions create`](https://docs.gumloop.com/cli/sessions) to run an agent with tools and conversation state.