LiveBench 2.0 Data Analysis
Professional Work · 2026-10-14
LiveBench 2.0 Data Analysis suite: Single-answer data analysis questions (consecutive events, table join, table reformat); 150 questions. LiveBench 2.0 (Abacus AI, released October 14, 2026) replaced the v1 question-answer categories with seven suites, five of them agentic; each suite is one run per model at its highest supported effort, a task the run could not measure scores 0, and a model without an input a suite needs (a text-only model on Browser) scores 0 there. Not comparable with LiveBench v1.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 85.2 |
| GPT-6 Astra | 83.0 |
| GPT-6.1 Sol | 82.7 |
| GPT-5.5 | 81.6 |
| Muse Wonder | 81.6 |
| GPT-6 Sol | 81.2 |
| Claude Fable 5 | 80.5 |
| Claude Fable 5.1 | 80.3 |
| Claude Opus 5.5 | 80.3 |
| GPT-5.6 Sol | 79.8 |
| Muse Spark 1.3 | 79.6 |
| DeepSeek V4 Flash Vision Exp | 79.5 |
| DeepSeek V4.1 Flash 1017 | 79.5 |
| DeepSeek V4.1 Flash | 79.3 |
| GPT-5.6 Terra | 79.3 |
| Kimi K3.1 Pro | 79.1 |
| Kimi K3 | 78.7 |
| Gemini 3.1 Pro Preview | 78.5 |
| GPT-5.6 Luna | 78.0 |
| Grok 4.7 | 76.9 |
| Qwen3.8 27B | 76.6 |
| Mistral Large 4 Preview | 76.5 |
| Muse Spark 1.2 | 76.5 |
| GLM-5.3 Flash | 76.4 |
| MiniMax M3 | 76.2 |
| Sonnet 4.6 | 76.1 |
| Grok 4.6 | 75.3 |
| Claude Opus 5 | 74.6 |
| Qwen3.8-Flash-Next | 74.2 |
| GPT-6 Luna | 73.4 |
| Grok 4.5 | 73.0 |
| Inkling | 72.8 |
| MiMo-V2.6-Pro | 72.3 |
| Sonnet 5 | 71.7 |
| MiniMax M3.1 Flash | 71.4 |
| GLM-5.3 | 70.2 |
| DeepSeek-V4-Flash-0731 | 68.0 |
| Gemini 3.7 Flash | 68.0 |
| Beam 1.0 501B A23B | 66.4 |
| Opus 4.8 | 66.0 |