add-driver

To add a new LLM provider integration, you can scaffold the necessary boilerplate code automatically. This generates sync and async driver classes, registers the provider, and configures the required settings and environment variables for seamless integration.

Scaffolding an LLM driver generates sync and async driver classes, registration entries, environment templates, package exports, setup.py extras, discovery integration, and models.dev pricing configurations to ensure the new provider is fully functional.

Setting up environment variables and package exports for a new LLM provider is handled automatically during the scaffolding process. The tool generates the necessary environment templates and updates package exports and setup.py extras.

Yes, you can generate both sync and async driver classes for an LLM provider. The scaffolding process creates both class types by default to ensure the new integration supports synchronous and asynchronous operations.

To configure models.dev pricing for a new LLM driver, the scaffolding process automatically generates the required pricing configurations. This ensures the newly integrated provider supports accurate pricing calculations immediately.

The best way to streamline LLM driver development is by automating the scaffolding of provider integrations. This eliminates manual boilerplate creation by generating driver classes, settings, and discovery integration automatically for each new provider.