TimeSage-MT - Decision Making: leaderboard
TimeSage-MT (2026) evaluates LLM agents in multi-turn time-series analysis dialogues with verifiable, code-derived answers, here under its Code-Enabled Reasoning setting in which the model writes Python that a sandbox executes. This board is the decision-making dimension on the 60 L4 tasks, a GPT-5.4 judge score in percent for the validity, reasoning and completeness of the final decision; higher is better.
Metric: Decision making (%): GPT-5.4 judge score of the final structured decision on the 60 L4 tasks (validity, reasoning quality, completeness), 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: Not forecast. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MiniMax-M2.7 | 64.08 | #518 |
| 2 | Qwen 3.5 122B A10B | 63.93 | #331 |
| 3 | GLM-5.1 | 62.88 | #179 |
| 4 | GPT-5.3 Codex | 59.45 | #116 |
| 5 | Claude Sonnet 4.6 | 59.38 | #161 |
| 6 | Gemini 3 Flash (Preview) | 52.69 | #144 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=timesage-mt-decision-making · How It Works · Data refreshed daily, snapshot 2026-11-13.