Recursive 发布开源编程模型 Flint：40B MoE 架构，激活参数仅 2.5B
Recursive Releases Flint, an Open-Weight 40B MoE Local Coding Model for Agents with 2.5B Active Params
Recursive 开源了 Flint，40B 总参数只激活 2.5B，SWE-bench 拿到 74.2%，Mac 上能跑到 125 token 每秒，本地跑编程 Agent 可以试试。
Recursive 发布了 Apache 2.0 许可的本地编程模型 Flint，采用 MoE 架构，总参数 40B，每次推理仅激活 2.5B 参数。官方报告其在 SWE-bench Verified 上取得 74.2% 的成绩。在 M4 Pro MacBook Pro 上运行速度超过每秒 125 个 token，可完全本地部署供编程智能体使用。
Recursive Releases Flint, an Open-Weight 40B MoE Local Coding Model for Agents with 2.5B Active Params
Recursive has released Flint, an Apache 2.0 local coding model with 40B total and 2.5B active parameters. It reports 74.2% on SWE-bench Verified and 125+ tokens per second on an M4 Pro MacBook Pro. The post Recursive Releases Flint, an Open-Weight 40B MoE Local Coding Model for Agents with 2.5B Active Params appeared first on MarkTechPost .
