AI Briefing
KO

27B flips the script

·2026.04.22 22:14

Key point

Qwen has released Qwen3.6-27B, an open-source model in the 27B class.

Details

Qwen has released Qwen3.6-27B. It is a 27B dense model, but they emphasize that its coding performance is flagship-level.

The key changes are as follows.

  • Strengthened agentic coding performance: surpasses Qwen3.5-397B-A17B on major coding benchmarks
  • Improved text and multimodal reasoning performance
  • Support for thinking / non-thinking modes
  • Fully open under the Apache 2.0 license

The official blog, Qwen Studio, GitHub, and Hugging Face model page were also released together. This is a launch aimed at delivering bigger results with a smaller model.

multimodalqwenopen-source-llmmodel-releasecoding-modelapache-2-0
Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Related

RedditNext GenAi Releases Rebekah 1.8B, an Open-Source Model for EdgeRedditSenseNova-U1.5-8B-MoT Released: Unified Text-to-Image and Understanding Model Initialized from Qwen3GeekNewsIFM Releases 'K2 Horizon', a Suite of 6 AI Models Exposing the Entire Training ProcessHacker NewsApple Releases LensVLM-9B for Processing Compressed Text ImagesTLDR AIQwen Releases 'RecreationWorld' for Training AI Agents to Rebuild AppsHacker NewsPrivatemode Benchmarks GLM-5.3-Flash as Vision-Capable Decision Model with Parity to Jev but Higher Cost