Alibaba Open Sources Qwen3.8-Flash: Undercuts DeepSeek, Runs 1M Context on 4090
量子位 · wechat · 2026-08-28
Alibaba Qwen released and open-sourced Qwen3.8-Flash, a 125B MoE model activating 6B parameters per token. It natively supports 262k context, extendable to 1M tokens.
Key Highlights:
- Performance: Topped Hugging Face trends upon release, outperforming Qwen3.7Plus in general, math, and coding.
- Pricing: Input costs 0.8 RMB per million tokens (output 2.7 RMB), roughly one-third the price of DeepSeek-V4-Flash.
- Deployment: Capable of running data-center-scale long context on consumer-grade 4090 GPUs.
- Benchmarks: Completed multi-platform copywriting in 2 minutes and processed 10k-word meeting minutes into action plans in 4 minutes during office scenario tests.
More from Infra
- SK Hynix ties new US plant to looser chip equipment exports to China, sources say — pstAsiatech · 2026-10-15
- Perplexity's new embedding models: index with a 9B model, search with 0.6B — antoine_chaffin · 2026-10-15
- Unsloth ships GGUF quants for Nemotron 3.5: Bolt-74B runs in 24GB RAM — danielhanchen · 2026-10-15
- vLLM-Omni ships Day-0 serving for FLUX 3 Image Klein with BF16, FP8, NVFP4 support — vllm_project · 2026-10-15
- NVIDIA's Nemotron 3.5 Super VL lands training and inference support on Applied Compute — ypatil125 · 2026-10-15
- CoreWeave bond yield tops 16% as loans outlast customer contracts by ~2 years — tengyanAI · 2026-10-15