Skip to content

Data Research and Analysis with Complex Operations

Agentic data-analysis tasks scored against per-task rubrics at a 980K-token budget.

6 models

Higher is better.

  1. 1Claude Opus 5Anthropic88.6
  2. 2Claude Sonnet 5.5Anthropic87.0
  3. 3Ling 3.1 FlashInclusionAI85.5
  4. 4Step 5 PreviewStepFun83.3
  5. 5Hy4 previewTencent77.2
  6. 6Ling 3.0 FlashInclusionAI70.4
Higher is better. Bar colour: blue proprietary, orange open weights, green other.

About this benchmark

Metric
Normalized rubric score
Tasks
Agentic data research and analysis tasks
Authors
Anthropic (2026)

Scores are as reported by model developers or evaluators and may use different settings (for example reasoning effort or tools). Compare with care.