Unverified50% confidenceOpinionExact time
The "Qwen3.8-27B" in the title is likely a memory error or typo by the poster, as the commonly seen Qwen specifications in the community are 32B, 72B, etc.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/25/2026
First Seen
Valid until: 11/23/2026
Sources
Qwen模型登顶HuggingFace热度榜前四背后的深层逻辑
Related Entities
Related Claims
UnverifiedQwen 3.6 35B A3B has a clear advantage over Gemma 4 and Qwen 3.5 in long-context memory benchmarks.71% similarUnverifiedQwen3.6 27B scored 77.2 on SWE-bench Verified.70% similarUnverifiedSome community users have speculated that Qwen may skip the 35B specification and directly release a MoE model at the 122B scale.70% similarUnverifiedQwen3.6 27B Q6 quantized version scored 670 points, ranking third.69% similarUnverifiedQwen3旗舰版（235B）的参数量仅为DeepSeek旗舰版（671B）的三分之一左右69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/800763API
curl https://kongchang.com/api/v1/knowledge/claims/800763MCP
get_claim(id=800763)