Qwen3.6-35B-A3B - The open sparse MoE model for agentic coding
Qwen3.6-35B-A3B is a highly efficient open-source MoE model with 35B total and just 3B active parameters. It delivers frontier-level agentic coding and multimodal reasoning, rivaling much larger dense models. Apache 2.0 licensed and available now.
Replies
Flowtica Scribe
Hi everyone!
Qwen3.6-35B-A3B is the first open-weight release in the Qwen3.6 family.
A 35B total / 3B active MoE, Apache 2.0, strong agentic coding, native multimodal reasoning, and dual thinking / non-thinking modes is a very serious package for an open release. The efficiency story is also part of the hook here: this is a sparse model trying to compete far above its active size.
And importantly, the surrounding workflow is already there. @OpenClaw , @Claude Code , Qwen Code, API path, self-hosting path — this is positioned as something people can actually build with right away.