TimeSage-MT - Factual Verification: leaderboard

Metric: Factual verification (%): rule-based match of data-grounded key facts (5% relative tolerance, key-finding coverage), averaged over the applicable tasks; Code-Enabled Reasoning setting (the model writes Python, a sandbox runs it, the model answers from stdout; 240 multi-turn time-series analysis tasks over 8 domains), temperature 0, reasoning effort high where supported and medium for GPT-5.3-Codex; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 6 models tracked.

Top models

#ModelScore
1Claude Sonnet 4.629.94
2GLM-5.128.79
3Qwen 3.5 122B A10B22.77
4MiniMax-M2.722.28
5GPT-5.3 Codex21.18
6Gemini 3 Flash (Preview)16.66

Interactive version: theaggregate.ai/benchmark?slug=timesage-mt-factual-verification · How It Works · Data refreshed daily, snapshot 2026-09-29.