Unverified50% confidenceBenchmarkExact time
The current top model's comprehensive test score is 58, 24 points above Qianwen 3-27B's 34.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/15/2026
First Seen
Valid until: 10/29/2026
Sources
Qwen3-27B Outperforms Last Year's Top Model — The Era of Small Models Has Truly Arrived
bilibiliAI大圆镜10/15/2026
Related Entities
Related Claims
UnverifiedThe third-party evaluation organization AI Mocha gave Qianwen 3-27B a score of 34 in a comprehensive test of math, code, and research abilities, while the champion of a year earlier, GPT-4o, scored only 23.77% similarUnverifiedThe model gave the correct answer of 34 in a multimodal image recognition test counting pencils, at a speed of 55 tokens per second at that time.74% similarUnverifiedThe locally run Gemma 26B achieved an accuracy of only 49.2% across 453 scored predictions, while its paired always-predict-up baseline had an accuracy of 53.4%.72% similarUnverifiedIn the experiment, the 27B model scored 61.0% on LiveMath when skills were written by the 4B model, whereas it scored 56.3% when it wrote the skills itself72% similarUnverified研究表明当模型参数规模达到3B以上时，1.58-bit模型在多项基准测试中表现可接近甚至匹配同等规模的全精度模型71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1039944API
curl https://kongchang.com/api/v1/knowledge/claims/1039944MCP
get_claim(id=1039944)