Unverified50% confidenceOpinionExact time
Qwen2.5-3B-Instruct、Phi-3.5-mini（3.8B）和Llama-3.2-3B-Instruct是2024年末在小参数规模上综合表现最优的三款模型
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/16/2026
First Seen
Valid until: 10/14/2026
Sources
Related Claims
UnverifiedQwen 3.6 35B A3B has a clear advantage over Gemma 4 and Qwen 3.5 in long-context memory benchmarks.75% similarVerifiedQwen3.6 35B-A3B是混合专家架构（MoE），总参数量35B，每次推理实际激活参数量约3B72% similarUnverifiedQwen3（千问3）在Ollama上提供0.6B/1.7B/4B/8B等多种规格71% similarUnverifiedQwen2.5-72B approaches the performance of Llama-3.1-405B across multiple benchmarks.70% similarUnverifiedIf Gemma 3 12B frequently makes errors on structured output, it is recommended to switch to the Qwen 2.5 series or Llama 3.1 8B/70B, as they have better native support for Function Calling.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/533258API
curl https://kongchang.com/api/v1/knowledge/claims/533258MCP
get_claim(id=533258)