Qwen3

Think Deeper or Act Faster

2.3K followers

2.3K followers

Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud. - QwenLM/Qwen3

Name: Qwen3
Rating: 5.0 (15 reviews)

Launch date

Qwen3.6-27BThe sweet-spot open dense model for coding agents

Launched on April 23rd, 2026

Qwen3.6-Max-PreviewThe flagship Qwen for agentic coding

Launched on April 22nd, 2026

Qwen3.6-35B-A3BThe open sparse MoE model for agentic coding

Launched on April 17th, 2026

Qwen3.6-PlusMultimodal AI optimized for real-world coding agents

Launched on April 3rd, 2026

Qwen3.5-OmniA native omni model for voice, video, and tools

Launched on March 31st, 2026

Qwen3.5 Small0.8B-9B native multimodal w/ more intelligence, less compute

Launched on March 3rd, 2026

Qwen3.5The 397B native multimodal agent with 17B active params

Launched on February 17th, 2026

Qwen3-TTSVoice design, cloning & 97ms streaming

Launched on January 23rd, 2026

Qwen-Image-2512SOTA open-source T2I model with even greater realism

Launched on January 1st, 2026

Qwen-Image-LayeredTurn flat images into multi-layer editable assets

Launched on December 22nd, 2025