Qwen Releases Qwen3.8-Max with 2.4 Trillion Parameters for Autonomous Coding

Qwen has announced Qwen3.8-Max, a 2.4 trillion parameter model designed for autonomous coding and long-running agentic workflows. The model can plan, write code, run tools, analyze results, and self-correct with minimal human intervention—demonstrated by the oh-my-cli project, where an agent developed a full application from an empty folder over 10+ days with a complete change log on GitHub.

API pricing is set at $2 per million input tokens, $6 for output, and $0.25 for cached input. Qwen plans to release open weights for Qwen3.8-Max and a compact Qwen3.8-27B variant next week, which would combine massive agentic capability with multimodality and open access.

Blog · Qwen Studio · API · GitHub