Scared to tears! Qwen 3.8 27B's ridiculous performance; can you believe this is only a 27B model? | Small Wave Mom's Radar Chart

Close to top-tier closed-source models, locally deployed Qwen3.8-27B generates mc in one sentence without modification

Running a 27B large model on 16GB VRAM? Qwen3.8-27B local test: 40T/s, this time it's truly legendary!

Qwen3.8-27B is officially open-sourced with full 27B parameters, and it can run on a single 4090

Testing qwen3.8 27b, beating deepseek-v4-flash without deep thinking mode turned on, please check the description

Watch how Qwen3.8 27B, with its light touch, leverages Opus4.6's heavy weight | The Shit Mountain Assessment

May's latest AI rankings, from gods to trash. Large language models, Agents, image/video generation - global AI rankings

Google Gemma 4 31B/26B locally tested like crazy! Benchmark results are impressive but crushed by Qwen 3.5-27B? 26B web generation crashes, 31B quantization collapses!

Who holds the future of AI?

The National Supercomputing Internet offers a token-plan, with the latest flash sale at 9.9 yuan per month, demonstrating successful execution of GLM5.2 configured with Claude!! Affordable prices, unlimited supply!! National GPU resources are abundant!!

Bravo! Qwen3.8 27b, this episode is amazing!

💀Bro is programming with a 0.5b model AI

How large of a model can your computer actually run?

Qwen3.6-35B-A3B: The first truly functional local open-source model! From 2,500 to 30,000, hands-on deployment across 3 different devices!

V100 running Qwen3.8 27B Qianwen nvfp4 with speed over 50 tokens, about half the speed of Pro 6000d. In the current environment of soaring graphics card prices, it is probably the most cost-effective AI operation graphics card.

Qwen3.8-27B disrecommend, meaningless for average users.

The relative ranking radar chart of Qwen 3.8 Max is actually a symmetrical figure | Lang Lang Ma's Radar Chart

39.5 Tokens per second! Qianwen 3.8 27B 2080Ti 22B single-card local deployment speedup.

[Share] Improve Qwen3.8 Thunder Big Thinking

A little fantasy: if only DeepSeek could surpass Claude Fable 5 one day, how nice would that be ｜ Lang Lang Ma Radar Chart

Personal Language Model Recommendation 2026 Latest Edition