← Back to live feed1 story
1
Alibaba Qwen3.8-27B Matches GPT-5.6 Luna, First Local Model to Hit Frontier Performance
Models45d agoThe Artificial Analysis Intelligence Index awarded a score of 52 to Alibaba's latest small-scale model, matching the benchmarks of DeepSeek V4-Pro. This is the first instance of an open-weight local model achieving frontier model capability, enabling the performance of the Qwen3.8-27B to be run on consumer hardware rather than massive cloud infrastructure.
The 27B parameter model has surged in adoption, with the Unsloth GGUF version recording 2.7 million downloads and reaching the number 2 trending spot on Hugging Face. Alibaba noted that the system is designed to run on laptops while maintaining the agentic planning and long-horizon task capabilities of the industry's largest frontier models.
Sign in to suggest edits
Key sources
- SOURCE@alibaba_qwen“A local 27B model scoring frontier performance!”x.com
- SUPPORT@alibaba_qwen“Strong enough to keep up with the frontier, light enough to run on your own laptop.”x.com
- SUPPORT@danielhanchen“the previous most-liked Unsloth GGUFs were Qwen3.6-35B-A3B with 1.54K likes and DeepSeek-R1 with 1.12K”x.com
- SOURCE@cline“We weren’t expecting this pace of local progress anywhere near this soon”x.com
- SUPPORT@xenovacom“can run locally... even in your browser with custom WebGPU kernels!”x.com