Qwen3.8-Flash-Next
The open-weight preview of Qwen4. Qwen3.8-Flash-Next is a 125B multimodal MoE with only 6B active parameters and a new architecture built around QSA, Gated Residual, N-gram embeddings, and Muon.
Reactions (11)
- Qwen3.8-Flash-Next – The open-weight preview of Qwen4external site
- Qwen3.8-Max – Qwen’s most capable model for coding and coworkexternal site
- Qwen3.6-27B – The sweet-spot open dense model for coding agentsexternal site
- Qwen3.6-Max-Preview – The flagship Qwen for agentic codingexternal site
- Qwen3.6-35B-A3B – The open sparse MoE model for agentic codingexternal site
- Qwen3.6-Plus – Multimodal AI optimized for real-world coding agentsexternal site
- Qwen3.5-Omni – A native omni model for voice, video, and toolsexternal site
- Qwen3.5 – The 397B native multimodal agent with 17B active paramsexternal site
- Qwen3-TTS – Voice design, cloning & 97ms streamingexternal site
- Qwen-Image-2512 – SOTA open-source T2I model with even greater realismexternal site
- Qwen-Image-Layered – Turn flat images into multi-layer editable assetsexternal site
Similar products
by description similarity
Capy-flash
github.coma 125B MoE model on two 16 GB consumer GPUs
Run a 105 GB AI model on a Mac that can't hold it. Qwen3.8-Flash-Next 4-bit, streamed from SSD on 16–64 GB Apple Silicon Macs. Fully native Swift, MLX and Metal. Offline, no Python. CLI, Swift library, and Ollama/OpenAI/Anthropic APIs. Works with Claude Code and Codex.
Qwen3.5 Small
huggingface.co0.8B-9B native multimodal w/ more intelligence, less compute
FlashQwen
github.comA from-scratch CUDA inference engine for Qwen3
Qwen3
github.comThink Deeper or Act Faster