HumanEval — Results

Rank	Model	Score
1	twil-8b	93.2
2	deepseek-v4-1-flash	79.4
3	minicpm-sala	0.951
4	kimi-k2-0905	0.945
5	gpt-5-1	0.94
6	claude-3-5-sonnet	0.937
7	gpt-5-2	0.936
8	gpt-5	0.934
9	gpt-5	0.934
10	kimi-k2-instruct	0.933
11	kimi-k2-instruct	0.933
12	qwen2-5-coder-32b-instruct	0.927
13	qwen2-5-coder-32b-instruct	0.926
14	sarvam-30b	0.921
15	claude-3-5-sonnet-v1	0.92
16	mistral-large-2	0.92
17	deepseek-v3-2	0.92
18	qwen2-5-vl-32b-instruct	0.915
19	gpt-4o	0.902
20	gemini-2-5-pro	0.9
21	deepseek-v3-1	0.898
22	granite-3-3-8b-base	0.897
23	granite-3-3-8b-instruct	0.897
24	nova-2-pro	0.89
25	llama-3-1-405b-instruct	0.89

H

HumanEval — Leaderboard

1 phaseActive

HumanEval Leaderboard

No self-reported entries.