humaneval-plus.

OCR benchmark

§ 01 · pass@1

pass@1.

Higher is better

#	Model	Score	Source
★	Qwen2.5-Coder-32B Fetched from CodeSOTA API on 2026-04-20	87.2	codesota-api
2	DeepSeek-V3 Fetched from CodeSOTA API on 2026-04-20	86.6	codesota-api
3	GPT-4o Fetched from CodeSOTA API on 2026-04-20	86	codesota-api
4	DeepSeek-Coder-V2 Fetched from CodeSOTA API on 2026-04-20	82.3	codesota-api
5	DeepSeek-Coder-33B Fetched from CodeSOTA API on 2026-04-20	75	codesota-api

§ Related · Explore