AI model benchmark leaderboard

Name: AI model benchmark leaderboard
Creator: LessWork Lab
License: https://creativecommons.org/licenses/by-sa/4.0/

Compare AI models using public benchmark scores, with links to the original source. LessWork Lab did not run these benchmark tests.

Benchmark source: LiveBenchBenchmark version date: 2026-06-25Source checked: 2026-10-10 11:01 (UTC+8)

Ranking method

Search models or providersMetric

69 / 69 entries

LiveBench 2026-06-25
Rank	Model	Provider	Overall
1	Claude Fable 5.1 Max Effort Expand category scores Reasoning 91.69 Coding 86.38 Agentic coding 66.06 Math 97.01 Data analysis 80.28 Language 89.50 Instruction following 72.99	Anthropic	83.41
2	Claude 5.5 Opus Thinking Max Effort Expand category scores Reasoning 92.15 Coding 89.25 Agentic coding 71.72 Math 97.08 Data analysis 80.31 Language 86.27 Instruction following 65.74	Anthropic	83.22
3	Claude Fable 5 Max Effort Expand category scores Reasoning 89.65 Coding 85.99 Agentic coding 62.17 Math 95.99 Data analysis 80.54 Language 90.68 Instruction following 75.77	Anthropic	82.97
4	GPT-6 Astra Max Effort Expand category scores Reasoning 92.65 Coding 80.36 Agentic coding 57.32 Math 96.81 Data analysis 82.97 Language 89.43 Instruction following 75.58	OpenAI	82.16
5	Claude 5.5 Opus Thinking xHigh Effort Expand category scores Reasoning 90.65 Coding 89.25 Agentic coding 65.35 Math 96.80 Data analysis 79.80 Language 85.50 Instruction following 67.05	Anthropic	82.06
6	GPT-6.1 Sol Max Effort Expand category scores Reasoning 92.63 Coding 80.36 Agentic coding 54.55 Math 96.83 Data analysis 82.66 Language 90.13 Instruction following 74.15	OpenAI	81.62
7	Muse Spark 1.3 xHigh Effort Expand category scores Reasoning 89.65 Coding 81.06 Agentic coding 64.09 Math 95.95 Data analysis 79.57 Language 82.79 Instruction following 78.00	Meta	81.59
8	DeepSeek V4.1 Flash Max Effort Expand category scores Reasoning 86.69 Coding 80.04 Agentic coding 77.27 Math 93.29 Data analysis 79.25 Language 81.19 Instruction following 70.03	DeepSeek	81.11
8	GPT-6.1 Sol xHigh Effort Expand category scores Reasoning 91.63 Coding 80.74 Agentic coding 56.77 Math 96.48 Data analysis 82.18 Language 88.64 Instruction following 71.33	OpenAI	81.11
10	GPT-5.6 Sol Max Effort Expand category scores Reasoning 91.65 Coding 83.94 Agentic coding 56.21 Math 96.20 Data analysis 79.84 Language 87.68 Instruction following 71.85	OpenAI	81.05
11	GPT-5.5 Thinking xHigh Effort Expand category scores Reasoning 89.65 Coding 82.15 Agentic coding 53.99 Math 95.86 Data analysis 81.58 Language 87.36 Instruction following 70.73	OpenAI	80.19
12	Claude 5 Opus Thinking Max Effort Expand category scores Reasoning 91.21 Coding 81.45 Agentic coding 65.20 Math 95.73 Data analysis 74.55 Language 88.69 Instruction following 63.77	Anthropic	80.08
13	Smaug Agentic Expand category scores Reasoning 90.27 Coding 82.47 Agentic coding 64.65 Math 83.92 Data analysis 79.90 Language 84.36 Instruction following 70.99	Abacus.AI	79.51
14	GPT-6 Sol Max Effort Expand category scores Reasoning 88.65 Coding 81.77 Agentic coding 52.88 Math 96.36 Data analysis 81.19 Language 85.30 Instruction following 68.57	OpenAI	79.25
15	Kimi K3 Expand category scores Reasoning 90.67 Coding 81.45 Agentic coding 62.17 Math 84.44 Data analysis 78.73 Language 85.53 Instruction following 71.36	Moonshot AI	79.19
16	Gemini 3.7 Flash High Expand category scores Reasoning 87.80 Coding 78.89 Agentic coding 58.28 Math 93.47 Data analysis 67.96 Language 85.45 Instruction following 79.93	Google	78.83
17	Qwen 3.8 Max Expand category scores Reasoning 88.21 Coding 72.87 Agentic coding 64.65 Math 91.31 Data analysis 78.41 Language 79.69 Instruction following 74.08	Alibaba	78.46
18	Grok 4.6 xHigh Expand category scores Reasoning 90.51 Coding 76.78 Agentic coding 57.02 Math 92.57 Data analysis 73.86 Language 83.70 Instruction following 71.87	xAI	78.04
19	GPT-5.4 Thinking xHigh Effort Expand category scores Reasoning 88.12 Coding 77.54 Agentic coding 53.84 Math 94.15 Data analysis 79.31 Language 82.63 Instruction following 70.22	OpenAI	77.97
20	Muse Spark 1.2 xHigh Effort Expand category scores Reasoning 90.00 Coding 77.54 Agentic coding 57.58 Math 91.20 Data analysis 76.46 Language 78.57 Instruction following 74.33	Meta	77.95
21	GPT-5.6 Terra Max Effort Expand category scores Reasoning 90.63 Coding 78.25 Agentic coding 54.95 Math 94.91 Data analysis 79.31 Language 82.89 Instruction following 64.62	OpenAI	77.94
22	Claude Sonnet 5.5 xHigh Effort Expand category scores Reasoning 86.81 Coding 88.87 Agentic coding 39.34 Math 96.69 Data analysis 78.62 Language 83.43 Instruction following 70.52	Anthropic	77.75
23	DeepSeek V4 Pro 0813 Expand category scores Reasoning 85.84 Coding 77.16 Agentic coding 54.95 Math 95.09 Data analysis 79.24 Language 82.07 Instruction following 67.70	DeepSeek	77.44
23	Smaug Flash Expand category scores Reasoning 86.23 Coding 77.10 Agentic coding 61.06 Math 90.06 Data analysis 79.12 Language 79.21 Instruction following 69.26	Abacus.AI	77.44
25	Grok 4.7 xHigh Expand category scores Reasoning 82.65 Coding 77.16 Agentic coding 53.99 Math 95.68 Data analysis 76.88 Language 80.14 Instruction following 75.28	xAI	77.40
26	Gemini 3.1 Pro Preview High Expand category scores Reasoning 84.00 Coding 76.45 Agentic coding 44.14 Math 91.04 Data analysis 78.54 Language 85.38 Instruction following 79.10	Google	76.95
27	Smaug Mini Expand category scores Reasoning 82.81 Coding 75.43 Agentic coding 60.81 Math 89.74 Data analysis 78.81 Language 77.00 Instruction following 73.92	Abacus.AI	76.93
28	DeepSeek V4 Flash Vision Exp Expand category scores Reasoning 85.40 Coding 68.20 Agentic coding 65.10 Math 87.81 Data analysis 79.48 Language 80.36 Instruction following 70.96	DeepSeek	76.76
29	Claude 4.7 Opus Thinking xHigh Effort Expand category scores Reasoning 87.19 Coding 82.09 Agentic coding 50.66 Math 92.85 Data analysis 78.26 Language 77.91 Instruction following 66.74	Anthropic	76.53
30	Claude 4.8 Opus Thinking Max Effort Expand category scores Reasoning 89.19 Coding 81.83 Agentic coding 50.50 Math 94.32 Data analysis 66.03 Language 79.66 Instruction following 72.03	Anthropic	76.22
31	Qwen 3.8 Flash Next Expand category scores Reasoning 87.38 Coding 72.55 Agentic coding 61.62 Math 85.82 Data analysis 74.24 Language 74.64 Instruction following 77.11	Alibaba	76.19
32	GLM-5.3 Expand category scores Reasoning 85.80 Coding 78.95 Agentic coding 60.91 Math 87.90 Data analysis 70.24 Language 79.86 Instruction following 69.30	Z.AI	76.14
33	Claude Sonnet 5 xHigh Effort Expand category scores Reasoning 88.69 Coding 80.68 Agentic coding 59.39 Math 92.94 Data analysis 71.74 Language 74.97 Instruction following 63.86	Anthropic	76.04
34	Gemini 3.8 Flash High Expand category scores Reasoning 89.29 Coding 72.49 Agentic coding 54.24 Math 91.56 Data analysis 54.01 Language 87.79 Instruction following 81.41	Google	75.83
35	Grok 4.5 Expand category scores Reasoning 87.17 Coding 68.59 Agentic coding 56.46 Math 90.82 Data analysis 73.04 Language 82.80 Instruction following 71.53	xAI	75.77
36	Claude Sonnet 5.5 Max Effort Expand category scores Reasoning 91.63 Coding 91.37 Agentic coding 56.31 Math 96.13 Data analysis 59.49 Language 77.98 Instruction following 56.79	Anthropic	75.67
37	Muse Spark 1.1 xHigh Effort Expand category scores Reasoning 87.73 Coding 77.16 Agentic coding 58.54 Math 87.14 Data analysis 72.55 Language 74.34 Instruction following 69.64	Meta	75.30
38	Qwen3.8 27B Expand category scores Reasoning 80.03 Coding 75.69 Agentic coding 61.36 Math 86.21 Data analysis 76.59 Language 74.35 Instruction following 72.66	Alibaba	75.27
39	Gemini 3.5 Flash High Expand category scores Reasoning 82.00 Coding 78.18 Agentic coding 48.99 Math 88.24 Data analysis 64.86 Language 84.58 Instruction following 75.60	Google	74.64
40	GPT-5.2 High Expand category scores Reasoning 83.21 Coding 76.07 Agentic coding 50.25 Math 93.17 Data analysis 78.16 Language 79.81 Instruction following 61.77	OpenAI	74.63
41	Claude 4.6 Opus Thinking High Effort Expand category scores Reasoning 88.67 Coding 78.18 Agentic coding 48.99 Math 89.32 Data analysis 69.89 Language 83.27 Instruction following 63.31	Anthropic	74.52
42	DeepSeek V4 Flash 0731 Expand category scores Reasoning 86.63 Coding 74.98 Agentic coding 46.77 Math 86.79 Data analysis 79.33 Language 79.17 Instruction following 65.52	DeepSeek	74.17
43	GPT-5.2 Codex Expand category scores Reasoning 77.71 Coding 83.62 Agentic coding 49.39 Math 88.77 Data analysis 78.20 Language 73.68 Instruction following 66.45	OpenAI	73.98
44	Gemini 3.6 Flash High Expand category scores Reasoning 85.15 Coding 77.86 Agentic coding 43.43 Math 86.40 Data analysis 63.00 Language 83.90 Instruction following 75.37	Google	73.59
45	GPT-5.6 Luna Max Effort Expand category scores Reasoning 85.64 Coding 82.91 Agentic coding 48.43 Math 87.20 Data analysis 78.03 Language 72.57 Instruction following 60.12	OpenAI	73.56
46	GLM-5.2 Expand category scores Reasoning 78.63 Coding 79.65 Agentic coding 51.77 Math 89.78 Data analysis 73.74 Language 76.24 Instruction following 62.29	Z.AI	73.16
47	Qwen 3.7 Max Expand category scores Reasoning 83.34 Coding 74.22 Agentic coding 43.59 Math 85.25 Data analysis 71.79 Language 79.74 Instruction following 74.04	Alibaba	73.14
48	Claude 4.6 Sonnet Thinking Medium Effort Expand category scores Reasoning 84.77 Coding 79.27 Agentic coding 42.63 Math 86.99 Data analysis 77.95 Language 76.10 Instruction following 63.22	Anthropic	72.99
49	Claude 4.5 Opus Thinking High Effort Expand category scores Reasoning 80.09 Coding 79.65 Agentic coding 39.70 Math 90.39 Data analysis 74.44 Language 81.26 Instruction following 62.55	Anthropic	72.58
50	Claude Haiku 5.5 xHigh Effort Expand category scores Reasoning 81.23 Coding 76.39 Agentic coding 51.36 Math 92.75 Data analysis 72.75 Language 63.52 Instruction following 66.47	Anthropic	72.07
51	GPT-6 Luna Max Effort Expand category scores Reasoning 81.77 Coding 78.95 Agentic coding 51.21 Math 89.12 Data analysis 73.37 Language 73.83 Instruction following 55.93	OpenAI	72.03
52	Inkling xHigh Effort Expand category scores Reasoning 78.35 Coding 71.02 Agentic coding 49.39 Math 88.36 Data analysis 72.78 Language 73.46 Instruction following 70.10	Thinking Machines	71.92
53	Mistral Large 4 High Expand category scores Reasoning 83.94 Coding 77.16 Agentic coding 57.17 Math 93.58 Data analysis 76.54 Language 49.64 Instruction following 64.82	Mistral AI	71.84
54	GLM-5.3 Flash Expand category scores Reasoning 77.64 Coding 78.95 Agentic coding 56.77 Math 81.24 Data analysis 76.40 Language 77.31 Instruction following 52.82	Z.AI	71.59
55	deepseek-v4-pro Display details pending verification Expand category scores Reasoning 82.69 Coding 69.99 Agentic coding 42.63 Math 90.68 Data analysis 74.54 Language 78.13 Instruction following 62.35	Not provided	71.57
56	Kimi K2.6 Thinking Expand category scores Reasoning 79.38 Coding 78.57 Agentic coding 46.92 Math 84.28 Data analysis 65.13 Language 75.14 Instruction following 64.36	Moonshot AI	70.54
57	GPT-5.4 Nano xHigh Expand category scores Reasoning 81.10 Coding 70.84 Agentic coding 46.77 Math 90.98 Data analysis 67.64 Language 62.51 Instruction following 67.20	OpenAI	69.58
58	ox-alpha-max Expand category scores Reasoning 76.59 Coding 75.75 Agentic coding 52.63 Math 77.53 Data analysis 75.77 Language 66.14 Instruction following 60.25	Stealth	69.24
59	Qwen 3.6 Plus Expand category scores Reasoning 75.83 Coding 78.18 Agentic coding 41.36 Math 83.72 Data analysis 69.91 Language 74.99 Instruction following 58.34	Alibaba	68.91
60	Kimi K2.7 Code Expand category scores Reasoning 82.81 Coding 73.96 Agentic coding 45.66 Math 79.60 Data analysis 62.66 Language 77.91 Instruction following 56.29	Moonshot AI	68.41
61	Claude Haiku 5.5 Max Effort Expand category scores Reasoning 84.63 Coding 77.86 Agentic coding 57.17 Math 86.44 Data analysis 51.78 Language 62.63 Instruction following 54.66	Anthropic	67.88
62	Grok Build 0.1 Expand category scores Reasoning 76.37 Coding 65.39 Agentic coding 45.81 Math 78.43 Data analysis 70.79 Language 72.46 Instruction following 65.22	xAI	67.78
63	Nemotron 3 Ultra 550B A55B Expand category scores Reasoning 74.70 Coding 70.70 Agentic coding 38.74 Math 88.66 Data analysis 54.46 Language 70.81 Instruction following 73.45	NVIDIA	67.36
64	Minimax M3 Expand category scores Reasoning 74.48 Coding 68.20 Agentic coding 40.66 Math 76.95 Data analysis 76.17 Language 76.84 Instruction following 57.51	Minimax	67.26
65	GPT-5.4 Mini xHigh Expand category scores Reasoning 71.32 Coding 71.62 Agentic coding 41.67 Math 78.46 Data analysis 70.79 Language 70.95 Instruction following 59.80	OpenAI	66.37
66	deepseek-v4-flash Display details pending verification Expand category scores Reasoning 70.58 Coding 69.23 Agentic coding 37.63 Math 79.65 Data analysis 68.02 Language 70.12 Instruction following 63.14	Not provided	65.48
67	Qwen 3.6 27B Expand category scores Reasoning 70.28 Coding 71.78 Agentic coding 39.29 Math 79.87 Data analysis 70.43 Language 63.30 Instruction following 53.23	Alibaba	64.03
68	Gemini 3.5 Flash-Lite High Expand category scores Reasoning 60.19 Coding 76.07 Agentic coding 45.25 Math 73.74 Data analysis 53.25 Language 71.82 Instruction following 67.24	Google	63.94
69	Grok 4.3 Expand category scores Reasoning 70.82 Coding 69.93 Agentic coding 18.54 Math 84.34 Data analysis 55.77 Language 73.58 Instruction following 62.75	xAI	62.25

How rankings work

Scores are out of 100. The overall score is the equally weighted mean of 7 category averages. Benchmark version 2026-06-25 includes 23 tasks.

Includes every model configuration with complete numeric scores in this benchmark version. When display details are missing, the source CSV's model ID is shown without inferring a name or provider. An unlisted configuration or a missing score does not imply weaker capability.

Compare scores from the same source and benchmark version. Keep the source's model names and list different reasoning settings separately. Rank using the source's scores and precision, with ties sharing a rank. Overall scores follow the official method and are rounded to 2 decimals; category rankings use unrounded means. These rankings are not product recommendations and do not compare price, speed, or image capabilities.

The source-check date does not establish which benchmark version is the latest. Scores do not cover every task.

Source checks are scheduled every 3 days. The checked date changes only after verification succeeds, the update passes QA, and it is published.

This page uses official LiveBench data. Filtering, rankings, and translations are edited by LessWork Lab. Original authors: LiveBench contributors. The data is shared under CC BY-SA 4.0, with links to the official leaderboard, source CSV, and license.

View the original leaderboard · LiveBench CSV · CC BY-SA 4.0