Recursive's Flint: automated research builds open local coding model, lm_head at 4-bit
_rockt · x · 2026-10-22
Richard Socher introduces Flint, Recursive SI's first model: a fast local coding model built from scratch in under 4 months by their automated research system, the Eureka Machine. It claims best-in-class fast local coding at 125+ tokens/sec on MacBook Pro and 150+ on RTX Spark, up to 2x faster than comparable models, with open weights on Hugging Face. Thom Wolf highlights that the system quantized the lmhead — which everyone says to keep in high precision — to 4 bits and was actually right, scoring just 0.7 points below bf16.
More from Models
- YC F26 startup Latent reads model internals to catch hallucinations at 530x lower cost — ycombinator · 2026-10-22
- Perplexity's JOLT matches GRPO with 6.5x fewer training tokens, JOLT+ wins on all five benchmarks — perplexity_ai · 2026-10-22
- Recursive AI ships Flint, a 40B-A2.5B MoE claimed to match Qwen3.6-27B, weights on Hugging Face — ChengleiSi · 2026-10-22
- DeepSeek ships V4.1-Flash-1017, now the #1 open model on RSI-Inference Bench — scaling01 · 2026-10-22
- OpenAI's Cybersecurity Guardrails Block a Dev From Auditing His Own Patched Server — Next_Technology6361 · 2026-10-22
- TPN-006 hits Hugging Face: 7.7GB compressed model beats its 17GB Mistral teacher — markjeffrey · 2026-10-22