LiveCodeBench

LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code. It continuously collects new problems from programming contests (LeetCode, AtCoder, CodeForces) and evaluates four different scenarios: code generation, self-repair, code execution, and test output prediction. Problems are annotated with release dates to enable evaluation on unseen problems released after a model's training cutoff.

id livecodebench · max 1 · 75 models reported

#	Model	Score
1	DeepSeek-V4-Pro-Max DeepSeek · open	0.94
2	DeepSeek-V4-Flash-Max DeepSeek · open	0.92
3	DeepSeek-V4-Flash-0423 DeepSeek · open	0.88
4	Solar Pro 4 Upstage	0.88
5	DeepSeek-V3.2 DeepSeek · open	0.83
6	DeepSeek-V3.2 (Thinking) DeepSeek · open	0.83
7	MiniMax M2 MiniMax · open	0.83
8	LongCat-Flash-Thinking-2601 Meituan · open	0.83
9	Nemotron 3 Super (120B A12B) NVIDIA · open	0.81
10	Grok-3 Mini xAI	0.80
11	Grok 4 Fast xAI	0.80
12	Grok-3 xAI	0.79
13	Grok-4 Heavy xAI	0.79
14	LongCat-Flash-Thinking Meituan · open	0.79
15	Grok-4 xAI	0.79
16	MiniMax M2.1 MiniMax · open	0.78
17	Nova 2 Pro Amazon	0.75
18	DeepSeek-V3.2-Exp DeepSeek · open	0.74
19	DeepSeek-R1-0528 DeepSeek · open	0.73
20	GLM-4.5 Zhipu AI · open	0.73
21	Nemotron Nano 9B v2 NVIDIA · open	0.71
22	Nova 2 Lite Amazon	0.71
23	GLM-4.5-Air Zhipu AI · open	0.71
24	Qwen3 235B A22B Alibaba Cloud / Qwen Team · open	0.71
25	Gemini 2.5 Pro Preview 06-05 Google	0.69
26	Mercury 2 Inception	0.67
27	Llama 3.1 Nemotron Ultra 253B v1 NVIDIA · open	0.66
28	Qwen3 32B Alibaba Cloud / Qwen Team · open	0.66
29	MiniMax M1 80K MiniMax · open	0.65
30	Ministral 3 (14B Reasoning 2512) Mistral AI · open	0.65
31	Mistral Small 4 Mistral AI · open	0.64
32	QwQ-32B Alibaba Cloud / Qwen Team · open	0.63
33	Qwen3 30B A3B Alibaba Cloud / Qwen Team · open	0.63
34	MiniMax M1 40K MiniMax · open	0.62
35	Ministral 3 (8B Reasoning 2512) Mistral AI · open	0.62
36	DeepSeek R1 Distill Llama 70B DeepSeek · open	0.57
37	DeepSeek R1 Distill Qwen 32B DeepSeek · open	0.57
38	DeepSeek-V3.1 DeepSeek · open	0.56
39	Qwen2.5 72B Instruct Alibaba Cloud / Qwen Team · open	0.56
40	Min istral 3 (3B Reasoning 2512) Mistral AI · open	0.55
41	Phi 4 Reasoning Microsoft · open	0.54
42	Kimi K2-Instruct-0905 Moonshot AI · open	0.54
43	DeepSeek R1 Distill Qwen 14B DeepSeek · open	0.53
44	Phi 4 Reasoning Plus Microsoft · open	0.53
45	Magistral Small 2506 Mistral AI · open	0.51
46	Magistral Medium Mistral AI · open	0.50
47	DeepSeek R1 Zero DeepSeek · open	0.50
48	QwQ-32B-Preview Alibaba Cloud / Qwen Team · open	0.50
49	DeepSeek-V3 0324 DeepSeek · open	0.49
50	LongCat-Flash-Chat Meituan · open	0.48
51	Llama 4 Maverick Meta · open	0.43
52	DeepSeek R1 Distill Llama 8B DeepSeek · open	0.40
53	DeepSeek R1 Distill Qwen 7B DeepSeek · open	0.38
54	DeepSeek-V3 DeepSeek · open	0.38
55	Gemini 2.0 Flash Google	0.35
56	Mistral Large 3 (675B Base) Mistral AI · open	0.34
57	Mistral Large 3 (675B Instruct 2512 Eagle) Mistral AI · open	0.34
58	Mistral Large 3 (675B Instruct 2512 NVFP4) Mistral AI · open	0.34
59	Mistral Large 3 (675B Instruct 2512) Mistral AI · open	0.34
60	Gemini 2.5 Flash-Lite Google · open	0.34
61	Llama 4 Scout Meta · open	0.33
62	Qwen2.5-Coder 32B Instruct Alibaba Cloud / Qwen Team · open	0.31
63	Gemini Diffusion Google	0.31
64	Gemma 3 27B Google · open	0.30
65	Qwen2.5 7B Instruct Alibaba Cloud / Qwen Team · open	0.29
66	Qwen2 7B Instruct Alibaba Cloud / Qwen Team · open	0.27
67	Gemma 3 12B Google · open	0.25
68	Qwen2.5-Coder 7B Instruct Alibaba Cloud / Qwen Team · open	0.18
69	DeepSeek R1 Distill Qwen 1.5B DeepSeek · open	0.17
70	Gemma 3n E2B Instructed Google	0.13
71	Gemma 3n E2B Instructed LiteRT (Preview) Google · open	0.13
72	Gemma 3n E4B Instructed Google	0.13
73	Gemma 3n E4B Instructed LiteRT Preview Google · open	0.13
74	Gemma 3 4B Google · open	0.13
75	Gemma 3 1B Google · open	0.02