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.

LiveCodeBench Leaderboard

AllOpenProprietary

69 models

#	Model	Score	Size	Context	Cost	License
1	DeepSeek-V3.2 DeepSeek	0.833	685B	164K	$0.26 / $0.38
1	DeepSeek-V3.2 (Thinking) DeepSeek	0.833	685B	—	—
3	MiniMax M2 MiniMax	0.830	230B	1.0M	$0.30 / $1.20
4	LongCat-Flash-Thinking-2601 Meituan	0.828	560B	128K	$0.30 / $1.20
5	Nemotron 3 Super (120B A12B) NVIDIA	0.812	120B	262K	$0.10 / $0.50
6	Grok-3 Mini xAI	0.804	—	128K	$0.30 / $0.50
7	Grok 4 Fast xAI	0.800	—	2.0M	$0.20 / $0.50
8	Grok-3 xAI	0.794	—	128K	$3.00 / $15.00
8	Grok-4 Heavy xAI	0.794	—	—	—
8	LongCat-Flash-Thinking Meituan	0.794	560B	128K	$0.30 / $1.20
11	Grok-4 xAI	0.790	—	—	—
12	MiniMax M2.1 MiniMax	0.780	230B	1.0M	$0.30 / $1.20
13	DeepSeek-V3.2-Exp DeepSeek	0.741	685B	—	—
14	DeepSeek-R1-0528 DeepSeek	0.733	671B	131K	$0.50 / $2.15
15	GLM-4.5 Zhipu AI	0.729	355B	131K	$0.40 / $1.60
16	Nemotron Nano 9B v2 NVIDIA	0.711	9B	—	—
17	Qwen3 235B A22B Alibaba Cloud / Qwen Team	0.707	235B	128K	$0.10 / $0.10
17	GLM-4.5-Air Zhipu AI	0.707	106B	—	—
19	Gemini 2.5 Pro Preview 06-05 Google	0.690	—	1.0M	$1.25 / $10.00
20	Mercury 2 Inception	0.670	—	128K	$0.25 / $0.75
21	Llama 3.1 Nemotron Ultra 253B v1 NVIDIA	0.663	253B	—	—
22	Qwen3 32B Alibaba Cloud / Qwen Team	0.657	33B	128K	$0.10 / $0.30
23	MiniMax M1 80K MiniMax	0.650	456B	1.0M	$0.55 / $2.20
24	Ministral 3 (14B Reasoning 2512) Mistral AI	0.646	14B	262K	$0.20 / $0.20
25	Mistral Small 4 Mistral AI	0.636	119B	256K	$0.15 / $0.60
26	QwQ-32B Alibaba Cloud / Qwen Team	0.634	33B	—	—
27	Qwen3 30B A3B Alibaba Cloud / Qwen Team	0.626	31B	128K	$0.10 / $0.44
28	MiniMax M1 40K MiniMax	0.623	456B	—	—
29	Ministral 3 (8B Reasoning 2512) Mistral AI	0.616	8B	262K	$0.15 / $0.15
30	DeepSeek R1 Distill Llama 70B DeepSeek	0.575	71B	128K	$0.10 / $0.40
31	DeepSeek R1 Distill Qwen 32B DeepSeek	0.572	33B	128K	$0.12 / $0.18
32	DeepSeek-V3.1 DeepSeek	0.564	671B	164K	$0.27 / $1.00
33	Qwen2.5 72B Instruct Alibaba Cloud / Qwen Team	0.555	73B	131K	$0.35 / $0.40
34	Min istral 3 (3B Reasoning 2512) Mistral AI	0.548	3B	131K	$0.10 / $0.10
35	Phi 4 Reasoning Microsoft	0.538	14B	—	—
36	Kimi K2-Instruct-0905 Moonshot AI	0.537	1.0T	—	—
37	Phi 4 Reasoning Plus Microsoft	0.531	14B	—	—
37	DeepSeek R1 Distill Qwen 14B DeepSeek	0.531	15B	—	—
39	Magistral Small 2506 Mistral AI	0.513	24B	—	—
40	Magistral Medium Mistral AI	0.503	24B	—	—
41	DeepSeek R1 Zero DeepSeek	0.500	671B	—	—
41	QwQ-32B-Preview Alibaba Cloud / Qwen Team	0.500	33B	33K	$0.15 / $0.60
43	DeepSeek-V3 0324 DeepSeek	0.492	671B	164K	$0.28 / $1.14
44	LongCat-Flash-Chat Meituan	0.480	560B	128K	$0.30 / $1.20
45	Llama 4 Maverick Meta	0.434	400B	1.0M	$0.17 / $0.85
46	DeepSeek R1 Distill Llama 8B DeepSeek	0.396	8B	—	—
47	DeepSeek R1 Distill Qwen 7B DeepSeek	0.376	8B	—	—
47	DeepSeek-V3 DeepSeek	0.376	671B	131K	$0.27 / $1.10
49	Gemini 2.0 Flash Google	0.351	—	1.0M	$0.10 / $0.40
50	Mistral Large 3 (675B Base) Mistral AI	0.344	675B	—	—

1–50 of 69

Prev1/2Next