Dev rents B200 cluster to run open-weight Qwen3.8 2.4T, clones COD in one prompt
Fun-Meaning-6474 · reddit · 2026-08-19
After Qwen open-sourced the 2.4T Max weights, the author rented a B200 cluster and tested it: roughly 1.1M output tokens over 5 hours to recreate a Call of Duty-style shooter in a single prompt.
While almost nobody can run this model locally, open weights unlock real opportunities for the local AI community — quantization options could bring frontier-level intelligence closer to consumer hardware. The 27B sibling from the same release is very capable for its size, with many 3D games built from it already posted on X. The author is from the atomic.chat team, which offers official and community quants of the 27B for direct download in-app.
More from Models
- Researcher: GPT 5.6 Sol Ultra Beats Pro for Long-Horizon Hard Problems — arankomatsuzaki · 2026-08-24
- Google Criticized: Gemini 3.7 Still Missing From Its Own Jules Agent a Week Later — brandon_galang · 2026-08-24
- Qwen 27B 3.8 low quantization tested: Q3 XXS works well locally — jeremyckahn · 2026-08-24
- Users notice significant quality shift in GPT-5.6 output — haider1 · 2026-08-24
- Ramp Stats: Anthropic Opus 4.8 and Sonnet 4.6 Lead Usage — vista8 · 2026-08-24
- Tencent Releases UI-Mate-27B, a Desktop GUI Agent Model — tencent · 2026-08-24