poolside/Laguna-M.1 · Hugging Face - 225B-A23B
The poolside/Laguna-M.1 model is a new 225B parameter Mixture-of-Experts (MoE) model with 23B activated parameters per token, designed for agentic coding and long-horizon tasks. It features a large sparse MoE architecture with 256 experts and top-k=16 routing, global attention, and native reasoning support for interleaved thinking. Laguna M.1 demonstrates strong performance on agentic benchmarks, including SWE-bench Verified, SWE-bench Multilingual, SWE-bench Pro, and Terminal-Bench 2.0, and is released under the Apache 2.0 license. AI
IMPACT This model's strong performance on coding benchmarks could accelerate the development of more capable AI agents for software engineering tasks.