Topics / #llm-routing
LLM routing projects on GitHub
Choosing the best model for each incoming request.
As of , the most-starred LLM routing projects on GitHub are awesome-free-llm-apis (10.6K stars), council-of-high-intelligence (4.7K stars) and hermes-jev-skills (1.1K stars). GitHub.Locker tracks 5 LLM routing repositories with 17.2K stars combined. The fastest-growing on October 17 was awesome-free-llm-apis, with +126 stars in a day.
Hot right now
Most stars gained in a day · Oct 17
+126awesome-free-llm-apis
mnfst/awesome-free-llm-apis · ★ 10.6KA curated list of LLM APIs that offer permanent free tiers for text inference. Every entry links straight to the provider's API key page, so you can register and start calling models at no cost.
+11council-of-high-intelligence
0xNyk/council-of-high-intelligence · ★ 4.7KCouncil of High Intelligence runs structured multi-perspective deliberation on hard decisions by assigning a question to 18 distinct analytical personas (Aristotle, Socrates, Feynman, Linus Torvalds and others), forcing independent first positions, cross-examination, and a final verdict that preserves unresolved questions, dissent, kill criteria, and a concrete next step.
+3hermes-jev-skills
kerpopule/hermes-jev-skills · ★ 1.1KHermes Jev Skills gives an AI agent a cheap, fast 'second brain' for small routine decisions (model routing, memory filtering, skill selection, turn compaction, computer/browser use) so the expensive frontier model only does the writing. The decisions are handled by Jev, TypeSafe's calibrated decision model.
The all-time top
5 projects · by stars
- 10.6K
#1 · mnfst/awesome-free-llm-apisawesome-free-llm-apis
A curated list of LLM APIs that offer permanent free tiers for text inference. Every entry links straight to the provider's API key page, so you can register and start calling models at no cost.
- 4.7K
#2 · 0xNyk/council-of-high-intelligencecouncil-of-high-intelligence
Council of High Intelligence runs structured multi-perspective deliberation on hard decisions by assigning a question to 18 distinct analytical personas (Aristotle, Socrates, Feynman, Linus Torvalds and others), forcing independent first positions, cross-examination, and a final verdict that preserves unresolved questions, dissent, kill criteria, and a concrete next step.
- 1.1K
#3 · kerpopule/hermes-jev-skillshermes-jev-skills
Hermes Jev Skills gives an AI agent a cheap, fast 'second brain' for small routine decisions (model routing, memory filtering, skill selection, turn compaction, computer/browser use) so the expensive frontier model only does the writing. The decisions are handled by Jev, TypeSafe's calibrated decision model.
- 4

Intent-Router
763angel291592/Intent-Router · PythonIntent-Router is an 'intent compiler' skill for AI coding agents: it turns a vague request into a typed IntentSpec contract by looking up what it can, asking only what a human must decide, or halting when the request is contradictory, ambiguous, or underspecified.
- 5

agent-router
115nidhi-singh02/agent-router · TypeScriptAgent Router is a CLI that looks at your AI coding-agent subscriptions, filters them by quota and policy rules, asks a ranking service for the best pick, and then launches the chosen agent (Cursor, Claude Code, Codex, or OpenCode) in a Herdr pane with your task.
Related topics
Often tagged together