free-coding-models
A CLI and dashboard that benchmarks live latency and stability across hundreds of free coding LLMs from many providers and installs the fastest one straight into a coding tool.
free-coding-models pings every model across roughly two dozen free and free-tier LLM providers in parallel, showing live latency and a computed stability score that combines p95 latency, jitter, spike rate, and uptime, since average latency alone hides a model that occasionally spikes for several seconds. Once a model is chosen, it writes the selection directly into the user's coding tool configuration and opens the tool.
It offers an interactive terminal dashboard as its default interface, a browser-based dashboard with shareable filtered views, plugins that let a model be swapped mid-session, and a local OpenAI-compatible router daemon with automatic failover. It also runs real-completion speed tests, a short setup wizard, and enriches its model list with drift detection and a tier scale based on a coding benchmark, and works with tools such as OpenCode, Aider, Cline, and Copilot.
It is written in JavaScript and distributed through npm.