← Back to live feed1 story
Recursive released code and demos enabling an eight-agent research swarm to run locally on a single 48GB M4 Pro Mac using one shared instance of its 40B parameter Flint model. The setup builds on Recursive's Eureka Machine, an automated AI research system that helped build Flint from scratch in under four months. Flint, released under an Apache 2.0 license with a 262K token context window, achieved 54.9% on Terminal-Bench 2.1, 89.7% on SWE-bench Pro v2, and 74.2% on SWE-bench Verified.
Sign in to suggest
Key sources
- SOURCE@cong_ml“Eight agents share one local 40B Flint model on a 48 GB M4 Pro: writing code, testing ideas and building on whatever works.”x.com
- SUPPORT@richardsocher“Introducing Flint, @Recursive_SI’s first model: a fast local coding model built with our automated research system, the Eureka Machine.”x.com
- SUPPORT@tydsh“Strong performance. 54.9% on Terminal-Bench 2.1, 89.7% on SWE-bench Pro v2, and 74.2% on SWE-bench Verified.”x.com
- SUPPORT@spolen23“Flint did 110 tok/s vs 28 and nailed every one. Qwen thought for 7 minutes on one and never answered.”x.com
- SUPPORT@thom_wolf“an AI researcher looking at the lm_head (that *everyone* tells you to keep in high precision) deciding to quantize it to 4 bits AND being actually right (0.7pt from bf16) is definitely impressive.”x.com
- SUPPORT@roberttlange“NVFP4 incl. a 4-bit lm_head (12% of active weights), self-distilled to within 0.7 pts of BF16”x.com
- SUPPORT@altryne“Tim Shi from @Recursive_SI: "We actually didn't have any pre-training, post-training effort until like 1 month ago"”x.com
- SUPPORT@_rockt“Great news for doing science on autoresearch swarms at home.”x.com