Recursive's Flint: a 40B MoE local coding model built by AI in under 4 months
ChengleiSi · x · 2026-10-22
Richard Socher unveiled Flint, Recursive AI's first model: a fast local coding model (40B MoE, 2.5B active, Apache 2.0) built almost entirely by their automated research system Eureka Machine — data, training recipes, inference kernels — from scratch in under 4 months. It claims 125+ tokens/s on an M4 Pro MacBook, 150+ on RTX Spark, up to 2x faster than peers, with open weights on Hugging Face.
More from AGI Musings
- AI industry internalized 'we're not cool' and pivots consumer apps at 35-year-olds — wordgrammer · 2026-10-22
- Peter Diamandis: AI doing 9th-grade homework is a travesty, using it to design spaceships is education — PeterDiamandis · 2026-10-22
- Physics says we are still very far from the limits of technology — burny_tech · 2026-10-22
- 'Zeno's Singularity': the name for gentle, low-slope recursive self-improvement — bradneuberg · 2026-10-22
- Researcher: Race for Speed Systematically Rewards Less Safe Model Development — davidmanheim · 2026-10-22
- Safety researcher: racing dynamics systematically undermine model safety, and it's not inevitable — Hugging Face Blog · 2026-10-22