Unverified50% confidenceBenchmarkExact time
早期的Codex模型在HumanEval上的通过率约为28.8%，而到2024-2025年GPT-4级别的模型已达到90%以上的通过率
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/24/2026
First Seen
Valid until: 11/22/2026
Sources
Related Entities
Related Claims
Unverified早期的Codex模型在HumanEval测试上的通过率约为28.8%74% similarUnverified自2023年GPT-4发布以来，AI在HumanEval等主流编程基准测试上的通过率从不足50%跃升至90%以上71% similarUnverifiedThe pass rate of top models in 2025 on the HumanEval benchmark has exceeded 90%.67% similarVerifiedPass rates of code generation models on benchmarks such as HumanEval jumped from 60% to over 90% between 2024 and 202567% similarUnverifiedA 2023 Stanford University study showed that multimodal models such as GPT-4V, when parsing flowcharts with nested conditional logic, had accuracy approximately 15–25 percentage points lower than equivalent textual descriptions.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/794928API
curl https://kongchang.com/api/v1/knowledge/claims/794928MCP
get_claim(id=794928)