Hackernews posts about Qwen 3

  1. Qwen3.8-Flash-Next (qwen.ai)
    704 points by tosh 25 days ago | 233 comments
  2. Qwen 3.8 27B available on Cerebras at 1500 tokens/s (inference-docs.cerebras.ai)
    691 points by altertable 17 days ago | 228 comments
  3. I gave Qwen 3.8 27B a reverse-engineering job and it finished in 30 minutes (www.xda-developers.com)
    369 points by raybb 28 days ago | 153 comments
  4. Qwen 3.8 Omni Flash (qwen.ai)
    341 points by jjcm 3 days ago | 137 comments
  5. Qwen 3.8-Flash-Next releasing tomorrow (125B a6B) (modelscope.cn)
    336 points by garo-pro 26 days ago | 173 comments
  6. Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses (quesma.com)
    286 points by stared 12 days ago | 139 comments
  7. Show HN: Running 104GB Qwen3.8-Flash-Next on 48GB Mac with at ~12 tok/s (github.com)
    240 points by carloslfu 19 days ago | 118 comments
  8. Qwen 3.8 follows GPT-5.5 Pro reasoning prefills (gist.github.com)
    236 points by wsxiaoys 11 days ago | 93 comments
  9. Run Qwen3.8 27B locally: real numbers from my Mac Studio (terminalbytes.com)
    140 points by speckx 23 days ago | 99 comments
  10. Shapelearn Qwen 3.8 27B (13.1 GB VRAM) (byteshape.com)
    104 points by syntaxing 3 days ago | 39 comments
  11. Show HN: Nari Qwen3-TTS and Qwen3-ASR – High accuracy, low latency and cost (narilabs.com)
    91 points by toebee 6 days ago | 31 comments
  12. Qwen 3.6 is now much easier to run locally on your Mac, thanks to JetBrains (www.neowin.net)
    39 points by bundie 27 days ago | 14 comments
  13. Qwen3.8-Flash-Next Technical Report [pdf] (github.com)
    35 points by Philpax 25 days ago | 1 comments
  14. Show HN: Swift-Qwen3.8-27B, -58.3% thinking, x1.95 speed, accuracy of xhigh (huggingface.co)
    30 points by kisjovan 4 days ago | 17 comments
  15. Qwen3.8-Max (www.qwencloud.com)
    29 points by Bluestein 18 days ago | 19 comments
  16. Run QWEN3.8 27B on 16gb Nvidia GPUs (github.com)
    28 points by Pragmata 3 days ago | 10 comments
  17. Qwen3.8-Flash-Next Intelligence, Performance and Price Analysis (artificialanalysis.ai)
    17 points by theanonymousone 24 days ago | 7 comments
  18. Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses (quesma.com)
    15 points by stared 25 days ago | 2 comments
  19. Qwen3.8-Max Checkpoint 0902 (www.qwencloud.com)
    14 points by fbrusch 18 days ago | 1 comments
  20. Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency (qwen.ai)
    12 points by _ache_ 25 days ago | discuss
  21. Qwen3.8-Flash-Next non-uniform quantization runs on 2 RTX3090s (huggingface.co)
    9 points by pfeifferj 6 days ago | 2 comments
  22. Unsloth/Qwen3.8-Flash-Next-GGUF (huggingface.co)
    9 points by Philpax 25 days ago | discuss
  23. I found why Qwen3.8-27B was 14× slower on my 4070 Super (github.com)
    8 points by logickkk1 28 days ago | discuss
  24. I Rented a 96 GB GPU and Took Uncensored Qwen3.8 From 44 to 125 tok/s (aseemshrey.com)
    7 points by LuD1161 28 days ago | 3 comments
  25. Qwen3.8-Max just got upgraded (twitter.com)
    7 points by mydreamof 18 days ago | 2 comments
  26. Show HN: Sageling - a local AI agent for Mac, Qwen 3.5 9B in-process via MLX (sageling.ai)
    7 points by kirillklimuk 16 days ago | 1 comments
  27. Qwen/Qwen3.8-Flash-Next (huggingface.co)
    7 points by Philpax 25 days ago | discuss
  28. GVS5H: Five Qwen3.8-27B Models Match Claude Fable 5 on LiveCodeBench Hard (github.com)
    6 points by OakNinja 8 days ago | discuss
  29. The Llama.cpp Fork That Enables Qwen 3.8 27B Large Contexts for 16GB VRAM GPU (github.com)
    5 points by dazhbog 20 days ago | 2 comments
  30. Qwen3.8 models outperforms Fable 5 Using GVS5H Harness (github.com)
    5 points by pggues 4 days ago | 1 comments
More