Inference
EN
LiveBench

LiveBench Data Analysis

Short questions on small tables: runs of growth, column matching, format changes. Questions from the 2026-06-25 release; LiveBench 2.0 keeps them.

ModelsModels with a result.80
Top resultBest result on this benchmark.85.2%Gemini 4 Argon
Top-3 spreadPoints from first to third.2.5 pts
UpdatedDate the source changed the results.14 Oct 2026

Result and cost

78 of 80 results

Results

80 results
#ModelResultThe score from the source.CostUS dollars to run one task.
01Gemini 4 Argonhigh effort85.2%$0.284
02GPT-6 Astramax effort83.0%$0.699
03GPT-6.1 Solmax effort82.7%$0.145
04GPT-6.1 Terramax effort82.3%$0.063
05Muse Wondermax effort81.6%$0.366
05GPT-5.5xhigh effort81.6%$0.405
07GPT-6 Solmax effort81.2%$0.199
08Claude Fable 5max effort80.5%$1.30
09Claude Opus 5.5max effort80.3%$0.815
09Claude Fable 5.1max effort80.3%$1.33
11Smaug Agentic79.9%$0.331
12GPT-5.6 Solmax effort79.8%$0.353
13Muse Spark 1.3xhigh effort79.6%$0.169
14DeepSeek V4 Flash Vision Exp79.5%$0.04
15DeepSeek V4 Flash 073179.3%$0.0077
15GPT-5.4xhigh effort79.3%$0.282
15GPT-5.6 Terramax effort79.3%$0.278
15DeepSeek V4.1 Flash 1017max effort79.3%$0.036
19DeepSeek V4 Pro 081379.2%$0.051
20Smaug Flash79.1%$0.0084
21Smaug Mini78.8%$0.027
22Kimi K378.7%$0.352
23Gemini 3.1 Pro Previewhigh effort78.5%$0.26
24Qwen3.8 Max78.4%$0.219
25Claude Opus 4.7xhigh effort78.3%$0.498
26GPT-5.2-Codex78.2%$0.199
26GPT-5.2high effort78.2%$0.166
28GPT-5.6 Lunamax effort78.0%$0.134
29Claude Sonnet 4.6 Thinking Automedium effort77.9%$0.326
30Grok 4.7xhigh effort76.9%$0.488
31Mistral Large 4.0high effort76.6%$0.063
31Qwen3.8-27B76.6%$0.02
33Mistral Large 4high effort76.5%$0.069
33Muse Spark 1.2xhigh effort76.5%$0.109
35GLM-5.3-Flash76.4%$0.024
36MiniMax M376.2%$0.029
37Claude Sonnet 4.6max effort76.1%$0.417
38Ox Alpha Max75.8%$0
39Grok 4.6xhigh effort75.3%$0.123
40Claude Opus 5max effort74.6%$0.746
41DeepSeek V4 Pro 042374.5%$0.043
42Claude Opus 4.5 Thinking 64Khigh effort74.4%$0.518
43Qwen3.8-Flash-Next74.2%$0.044
44GLM-5.273.7%$0.113
45GPT-6 Lunamax effort73.4%$0.017
46Grok 4.573.0%$0.115
47Inklingxhigh effort72.8%$0.17
48Muse Spark 1.1xhigh effort72.5%$0.087
49MiMo-V2.6-Prohigh effort72.3%$0.039
50Qwen3.7 Max71.8%$0.121
51Claude Sonnet 5xhigh effort71.7%$0.465
52MiniMax M3.1 Flashmax effort71.4%$0.012
53Grok Build 0.170.8%$0.0042
53GPT-5.4 minixhigh effort70.8%$0.197
55Qwen3.6-27B70.4%$0.088
56GLM-5.370.2%$0.373
57Qwen3.6 Plus69.9%$0.077
57Claude Opus 4.6 Thinking Autohigh effort69.9%$0.362
59DeepSeek V4 Flash 042368.0%$0.013
59Gemini 3.7 Flashhigh effort68.0%$0.167
61GPT-5.4 nanoxhigh effort67.6%$0.045
62Beam 1.0max effort66.4%$0.075
63Claude Opus 4.8max effort66.0%$0.646
64Nemotron Super VL-3.5 Nvfp4max effort65.7%$0.017
65Kimi K2.6 Thinking65.1%$0.125
65Smaug Nemotron 3.5 Nvfp4max effort65.1%$0.017
67Gemini 3.5 Flashhigh effort64.9%$0.246
68Nemotron Super VL-3.5max effort64.6%$0.019
69Step 5 Previewmax effort63.5%$0.064
70Gemini 3.6 Flashhigh effort63.0%$0.191
71Kimi K2.7 Code62.7%$0.089
72Muse Spark Openmax effort60.2%$0.045
73Claude Sonnet 5.5max effort59.5%$0.531
74Claude Haiku 4.5max effort59.3%—
75Grok 4.355.8%$0.035
76Gemini 3.8 Flashhigh effort54.0%$0.238
77Gemma 4 31Bmax effort53.3%$0.0046
78Gemini 3.5 Flash-Litehigh effort53.2%$0.045
79Claude Haiku 5.5max effort52.4%$0.03
80Nemotron 3.5 Lightning 30B A3B43.1%$0.014