Repolode
Submit
vava-nessa

vava-nessa/free-coding-models

A CLI and web dashboard that benchmarks free coding LLM models across providers and installs the chosen one into your coding tool.

View on GitHubWebsite
Stars
2,824
Forks
294
Open issues
4
Last commit
yesterday

What is it?

free-coding-models finds the fastest and most stable free coding LLM from a catalog of several hundred models across roughly two dozen providers. It pings every model in parallel, shows live latency, and computes a stability score from p95 latency, jitter, spike rate and uptime, so a model that is fast on average but spikes badly ranks lower.

Once you choose a model, the tool writes it into the configuration of a supported coding tool such as OpenCode, Goose, Aider, Crush or Cline, then launches that tool. The main interface is a terminal dashboard, with a browser dashboard, an OpenAI-compatible local router that fails over between models, and plugins for OpenCode and Pi also available.

It is distributed as an npm package that runs on Node.js 18+, and it stores API keys locally in a user config file.

Key features

  • 01Parallel live latency and stability benchmarks for free coding models
  • 02Writes the selected model into supported coding tools and launches them
  • 03Terminal UI with filters, sorting and a Smart Recommend wizard
  • 04Browser dashboard with shareable filtered views
  • 05Local OpenAI-compatible router with automatic failover
  • 06Plugins for OpenCode and Pi

Use cases

  • Picking a free model for daily coding without trial and error
  • Comparing provider latency and reliability before committing to one
  • Routing an editor or agent through one local endpoint that survives provider outages

Quick start

  1. 1

    Install the CLI globally with npm

    npm install -g free-coding-models
  2. 2

    Launch the interactive dashboard and paste your free API key when prompted

    free-coding-models

Things to know

  • Requires a free API key from one of the 24 providers to use most models.
  • Health probes consume provider quota, and providers are paused on 429 responses.
  • License is listed as NOASSERTION, so the project's own license is unclear.

Topics

#llm#cli#coding-assistant#benchmark#openai-compatible#terminal#ai#deepseek#free#free-ai

FAQ

What is free-coding-models?
A CLI and web dashboard that benchmarks free coding LLM models across providers and installs the chosen one into your coding tool. free-coding-models finds the fastest and most stable free coding LLM from a catalog of several hundred models across roughly two dozen providers. It pings every model in parallel, shows live latency, and computes a stability score from p95 latency, jitter, spike rate and uptime, so a model that is fast on average but spikes badly ranks lower.
Is free-coding-models free to use? What license does it have?
free-coding-models is open source under the Other license. Check the license terms before using it in commercial products.
Is free-coding-models still maintained?
Actively maintained. The last commit to free-coding-models was on Oct 19, 2026.
What can I use free-coding-models for?
Picking a free model for daily coding without trial and error Comparing provider latency and reliability before committing to one Routing an editor or agent through one local endpoint that survives provider outages
How do I install free-coding-models?
Install the CLI globally with npm: npm install -g free-coding-models Launch the interactive dashboard and paste your free API key when prompted: free-coding-models
Where can I get free-coding-models?
The source code and installation instructions are on GitHub: https://github.com/vava-nessa/free-coding-models

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects