StackMap
Subscribe

9router vs free-claude-code

Self-hosted gateway pointing any AI coding CLI (Claude Code, Codex, Cursor, Cline) at 40+ providers, with subscription→cheap→free auto-fallback and tool_result compression for 20-40% token savings. — versus — Provider-backed proxy that runs Claude Code, Codex or Pi on 25 cloud and local providers — fcc-* launchers, local Admin UI with validation, per-tier model routing, IDE/Discord/Telegram hookups.

The curated verdict

Both are self-hosted gateways pointing Claude Code/Codex-class CLIs at many providers. 9router leans on routing policy (subscription→cheap→free fallback, token compression); FCC leans on UX — launchers, validation UI, native model-picker integration, IDE configs.

9routerfree-claude-code
Stars22k40k
Forks3.8k6.6k
LanguageJavaScriptPython
LicenseMITMIT
Last activity6 days agotoday
Topicsgateway, codinggateway
Curated connections52

9router — the curator's take

The 'never hit a rate limit again' gateway: run it on localhost, point Claude Code / Codex / Cursor / Cline / Copilot at its OpenAI-compatible endpoint, and it round-robins your accounts and cascades Subscription → cheap ($0.2-0.6/1M) → free (Kiro, OpenCode) providers so a job never dies mid-flight, while RTK compresses tool_result payloads (git diff, grep, ls) to shave 20-40% of tokens. Best when you juggle several provider subscriptions/keys and keep exhausting quota. NOT for you if you want one stable premium model (free-tier fallbacks vary in quality/availability), if routing traffic through third-party free providers raises data-privacy concerns, or if you want savings from smarter code retrieval rather than payload compression. Overlaps heavily with lynkr — pick 9router for free/multi-account fallback breadth, lynkr for tool-stripping, semantic caching and local-model routing.

free-claude-code — the curator's take

40k stars because it nails onboarding: one installer, `fcc-server` + `fcc-claude`, paste an NVIDIA NIM key and Claude Code runs for free — the Admin UI validates providers, and the agents' NATIVE /model pickers see the whole FCC catalog. Tier routing is the power feature: send Opus traffic to Kimi, Sonnet to an OpenRouter free route, Haiku to a local LM Studio model. Local models (Ollama, llama.cpp, LM Studio) are first-class. NOT a quota-stacker — you pick a provider rather than multiplex free tiers with failover (that's freellmapi's job), and expect capability drift vs real Anthropic models on complex agentic work. Discord/Telegram remote sessions with voice-note transcription are a genuinely odd, genuinely useful extra.