Evaluate models with the CLI

Use the EvalRouter CLI to discover available evaluations and models, review a spending cap, run an evaluation, and export its results.

1. Install and run your first evaluation.
2. Create an account and workspace key.
3. Check the supported commands.
4. Read results and export reports.

Choose exact model routes and benchmark profiles from your workspace's catalog. A listing is not a promise that every model and profile are compatible. Review the quote before starting work.

The initial public offering is the command-line model-evaluation workflow. See availability and limits.

Use the model API reference for direct HTTP integrations.