Coding

LiveCodeBench Pro leaderboard

Name: benchgap: LLM benchmark scores, measured and estimated
Creator: benchgap

As of 2026-10-07, the highest measured score on LiveCodeBench Pro is 90.8% by Sakana Fugu-Ultra. 67 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#	Model	Score	Source
1	Claude Opus 5.5	100.0%	estimated ± 6.0 pp, low confidence
2	Claude Sonnet 5.5	98.0%	estimated ± 6.0 pp, low confidence
3	Claude Fable 5.1	97.9%	estimated ± 6.0 pp, low confidence
4	Claude Mythos 5	97.4%	estimated ± 6.0 pp, low confidence
5	Claude Fable 5	97.2%	estimated ± 6.0 pp, low confidence
6	Claude Opus 5	96.8%	estimated ± 6.0 pp, low confidence
7	Sakana Fugu-Ultra	90.8%	measured
8	Claude Opus 4.8	90.4%	estimated ± 6.0 pp, low confidence
9	Qwen3.8 Max	89.4%	estimated ± 6.0 pp, low confidence
10	Hy4 preview	88.0%	estimated ± 6.0 pp, low confidence
11	Beam	87.8%	estimated ± 6.0 pp, low confidence
12	Sakana Fugu	87.8%	measured
13	Ornith-1.5-397B	87.5%	estimated ± 6.0 pp, low confidence
14	GPT-5.4	87.5%	measured
15	Grok 4.5	87.3%	estimated ± 6.0 pp, low confidence
16	GPT-5.6 Sol	87.2%	estimated ± 6.0 pp, low confidence
17	Claude Opus 4.7 (Adaptive)	87.0%	estimated ± 6.0 pp, low confidence
18	GPT-5.6 Terra	86.3%	estimated ± 6.0 pp, low confidence
19	Qwen3.8-Omni-Flash	86.2%	estimated ± 6.0 pp, low confidence
20	Claude Sonnet 5	86.2%	estimated ± 6.0 pp, low confidence
21	GPT-5.6 Luna	85.8%	estimated ± 6.0 pp, low confidence
22	Qwen3.8-Flash-Next	85.6%	estimated ± 6.0 pp, low confidence
23	Ornith-1.0-397B	85.4%	estimated ± 6.0 pp, low confidence
24	GLM-5.2	85.3%	estimated ± 6.0 pp, low confidence
25	Qwen3.8-27B	85.0%	estimated ± 6.0 pp, low confidence
26	Muse Spark 1.1	84.9%	estimated ± 6.0 pp, low confidence
27	dots3-note Preview	84.5%	estimated ± 6.0 pp, low confidence
28	Qwen3.7 Max	84.2%	estimated ± 6.0 pp, low confidence
29	Atria Dawn Preview	83.4%	estimated ± 6.0 pp, low confidence
30	Ornith-1.5-35B-A3B	83.4%	estimated ± 6.0 pp, low confidence
31	Laguna S 2.1	83.3%	estimated ± 6.0 pp, low confidence
32	MiniMax M3	83.0%	estimated ± 6.0 pp, low confidence
33	Gemini 3.1 Pro	82.9%	measured
34	GPT-5.5	82.6%	estimated ± 6.0 pp, low confidence
35	Kimi K2.6	82.6%	estimated ± 6.0 pp, low confidence
36	GLM-5.1	82.5%	estimated ± 6.0 pp, low confidence
37	Qwen3.7 Plus	81.8%	estimated ± 6.0 pp, low confidence
38	Qwen 3.6 Max (preview)	81.6%	estimated ± 6.0 pp, low confidence
39	MiMo-V2.5-Pro	81.5%	estimated ± 6.0 pp, low confidence
40	Claude Opus 4.5	81.4%	estimated ± 6.0 pp, low confidence
41	GPT-5.3 Codex	81.2%	estimated ± 6.0 pp, low confidence
42	Ling 3.0 Flash	81.0%	estimated ± 6.0 pp, low confidence
43	Qwen3.6 Plus	81.0%	estimated ± 6.0 pp, low confidence
44	Step 3.7 Flash	80.8%	estimated ± 6.0 pp, low confidence
45	MiniMax M2.7	80.7%	estimated ± 6.0 pp, low confidence
46	MiMo-V2.5	80.6%	estimated ± 6.0 pp, low confidence
47	Inkling-Small	80.5%	estimated ± 6.0 pp, low confidence
48	GPT-5.2	80.2%	estimated ± 6.0 pp, low confidence
49	DeepSeek V4 Pro 0813	80.0%	estimated ± 6.0 pp, low confidence
50	Muse Spark	80.0%	measured
51	Gemini 3.5 Flash	79.8%	estimated ± 6.0 pp, low confidence
52	GLM-5	79.8%	estimated ± 6.0 pp, low confidence
53	Inkling	79.1%	estimated ± 6.0 pp, low confidence
54	Gemini 3.5 Flash-Lite	79.0%	estimated ± 6.0 pp, low confidence
55	Qwen3.6-27B	78.4%	estimated ± 6.0 pp, low confidence
56	MAI-Thinking-1	77.8%	estimated ± 6.0 pp, low confidence
57	DeepSeek V4 Flash 0731	77.7%	estimated ± 6.0 pp, low confidence
58	Muse Glimmer 30B	76.4%	estimated ± 6.0 pp, low confidence
59	Qwen3.5 397B	76.2%	estimated ± 6.0 pp, low confidence
60	Kimi K2.5	76.0%	estimated ± 6.0 pp, low confidence
61	Ornith-1.0-35B	75.7%	estimated ± 6.0 pp, low confidence
62	Qwen3.6-35B-A3B	74.9%	estimated ± 6.0 pp, low confidence
63	Laguna M.1	74.6%	estimated ± 6.0 pp, low confidence
64	Grok 4.20	74.2%	measured
65	Laguna XS 2.1	73.2%	estimated ± 6.0 pp, low confidence
66	Ornith-1.5-9B	73.1%	estimated ± 6.0 pp, low confidence
67	Laguna XS.2	71.9%	estimated ± 6.0 pp, low confidence
68	Claude Opus 4.6	70.7%	measured
69	Ornith-1.0-9B	68.6%	estimated ± 6.0 pp, low confidence
70	LongCat-Flash-Lite-Sparse	66.2%	estimated ± 6.0 pp, low confidence
71	Granite 4.2 30B	57.9%	estimated ± 6.0 pp, low confidence
72	LLaDA2.2-flash	54.0%	estimated ± 6.0 pp, low confidence
73	Granite 4.2 8B	38.3%	estimated ± 6.0 pp, low confidence
74	MiniCPM5-2B	30.4%	estimated ± 6.0 pp, low confidence
75	MiniCPM5-1B	22.7%	measured

LiveCodeBench Pro leaderboard

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General