StatABench - AB Test: leaderboard

Metric: Accuracy (%) on the 20 ab test questions of Stat-Closed; Stat-Closed questions (multiple choice, fill in the blank, decision making, practical application with datasets) answered with the SAToolKit statistical functions through the LangChain MCP baseline tool-use setting, temperature 0; direct matching for closed answers and an LLM judge for open-form answers; higher is better. Source: arxiv.org. Saturation forecast: Around June 2027. 9 models tracked.

Top models

#ModelScore
1DeepSeek V375
2Claude Sonnet 4.555
3GPT-5.155
4Qwen 2.5 72B Instruct50
5Qwen 2.5 7B Instruct50
6Qwen 3 8B (Non-reasoning)45
7Qwen 2.5 32B Instruct40
8Llama 3.1 8B Instruct35
9GPT-4o Mini35

Interactive version: theaggregate.ai/benchmark?slug=statabench-ab-test · How It Works · Data refreshed daily, snapshot 2026-09-29.