Launching Monday, November 16, 2026
00
:
00
:
00
:
00
days
hours
mins
secs

About

Ficelle runs on your machine and gives OpenAI-compatible clients one local API for free LLM tiers. It routes requests to upstream providers, handles failover and cooldowns, and shows routing activity in a browser dashboard. Paid fallback is blocked in code; provider quotas still apply. Ficelle requires Python 3.11+ and supports macOS, Linux and Windows. Inference runs with the upstream providers. Core is free and source-available under BUSL-1.1, with optional Pro features.