Unverified50% confidenceBenchmarkExact time
Qwen2.5-27B是阿里云通义实验室发布的270亿参数开源模型，在MMLU、HumanEval、GSM8K等基准上接近GPT-4早期版本水平
1
Sources
50%
Confidence
Long-term
Relevance
7/17/2026
First Seen
Sources
Related Claims
UnverifiedOpen-source models such as Qwen-7B and LLaMA-2-7B, after Agent Tuning, can achieve tool-calling accuracy at the level of GPT-3.5 or even close to GPT-477% similarUnverifiedOpen-source models such as Llama 3.1 (Meta), Mistral 7B/8x7B, and Qwen2.5-Coder (Alibaba Cloud) have reached practical performance levels approaching GPT-4 on code and configuration generation tasks.76% similarUnverifiedOpen-source models such as Meta's Llama 3.1 series, DeepSeek-V3, and Alibaba's Qwen2.5-Coder have significantly narrowed the gap with GPT-4-level models on code generation benchmarks such as HumanEval and SWE-bench.72% similarUnverifiedOn mainstream benchmarks such as MMLU, HumanEval, and GSM8K, open-source models have reached 85–95% of the level of early versions of GPT-4.72% similarVerifiedCurrently mainstream agent base models include closed-source models such as GPT-4o, Claude 3.5 Sonnet, and Gemini 1.5 Pro, as well as open-source models such as Llama 3, Qwen2.5, and DeepSeek71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/538898API
curl https://kongchang.com/api/v1/knowledge/claims/538898MCP
get_claim(id=538898)