DownloadQwen3.8-Flash-Next is Qwen’s experimental open-weight multimodal model previewing the architecture planned for Qwen4. It uses 125B language-model parameters with only 6B activated per token, supplemented by 51B n-gram embedding parameters and 4B for multi-token prediction. Its hybrid architecture combines Gated DeltaNet with Qwen Sparse Attention (QSA), which processes micro-blocks rather than individual tokens to reduce latency in long-context agent workloads. The model also introduces Gated Residual connections and scalable n-gram embeddings to improve efficiency while limiting inference overhead. It contains 512 MoE experts, activating 10 routed experts plus one shared expert per token. Qwen3.8-Flash-Next natively handles text, images, and video and supports a 262K-token context window extensible to 1M tokens. It targets coding, tool use, professional tasks, computer interaction, multimodal reasoning, and long-horizon agents, with configurable thinking modes and reasoning effort.
Features
- 125B parameters with only 6B activated per token
- 51B n-gram embedding and 4B multi-token prediction parameters
- Hybrid Gated DeltaNet and Qwen Sparse Attention architecture
- 512 experts with 10 routed plus one shared expert activated
- Native text, image, and video understanding
- 262K native context extensible up to 1M tokens
- Configurable thinking and reasoning-effort modes
- Optimized for coding, tool use, computer interaction, and agents
Project Samples
