running-the-proxy

You can manage rate limiting and budgets for OpenAI-compatible API keys by configuring virtual keys through a TOML configuration file. This allows you to assign per-key rate limits, track costs, and control model routing directly.

Routing requests to multiple LLM providers is handled by configuring model routing within your gateway settings. This proxy supports routing to 143 provider models, managed easily using virtual keys and a TOML configuration file.

You start the OpenAI-compatible proxy for cost tracking and gateway management by running the command `liter-llm api --config liter-llm-proxy.toml`. This initializes server, model, and key settings defined in your TOML file.

Yes, you can configure server settings, general configurations, models, and virtual keys using a TOML file for your LLM gateway. The proxy operates entirely through this TOML configuration file for settings and key management.

Yes, the gateway management proxy provides a RESTful API for control and data exchange. This allows you to programmatically interact with virtual keys, budgets, and cost tracking data alongside the TOML configuration file.