Unverified50% confidenceBenchmarkExact time
Benchmarked against Qwen3.8-27B, Agens Volundr 32B leads on LiveCodeBench v6 (+4.2), HumanEval (+4.3), AIME 2025 (+2.9), and MATH-500 (+1.6).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/5/2026
First Seen
Valid until: 1/3/2027
Sources
Related Entities
Related Claims
UnverifiedAlibaba Cloud's Qwen2.5-Coder-32B surpassed GPT-4o on mainstream code benchmarks including HumanEval and MBPP in evaluations conducted at the end of 2024.70% similarUnverifiedOfficial benchmark data shows that the new version of Qwen3 27B agent coding improved from 13.3 to 42.2, software engineering tasks improved from 49.3 to 79, and frontier agent tasks improved from 10.6 to 20.67% similarUnverifiedThe performance of Qwen2.5-Coder and DeepSeek-Coder-V2 on code benchmarks such as HumanEval and MBPP is already very close to that of top closed-source models such as Claude 3.5 Sonnet.65% similarUnverifiedThe dataset covers four open-weight models — Qwen2.5-7B, Mistral-7B-v0.3, Phi-3.5-mini, and Qwen2.5-1.5B — with 30 trajectories each.65% similarUnverifiedReddit社区的大规模基准测试基于BeeLlama.cpp v0.4.0，在Qwen 3.6 27B和Gemma 4 31B两款模型上累计跑了413组配置对比，其中Qwen模型238组、Gemma模型175组64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/977117API
curl https://kongchang.com/api/v1/knowledge/claims/977117MCP
get_claim(id=977117)