TemporalBench - T1: leaderboard
TemporalBench T1 family measuring MCQ accuracy on contextual time-series understanding across retail, process, causal, and clinical datasets.
Metric: Accuracy (%). Source: huggingface.co. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | TimeCopilot (deepseek-chat) | 50 | |
| 2 | Single LLM (gpt-4o) | 49.72 | |
| 3 | CAMEL (gpt-4o) | 49.44 | |
| 4 | Single LLM (deepseek-chat) | 49.44 | |
| 5 | MetaGPT (deepseek-chat) | 49.16 | |
| 6 | AgentScope (deepseek-chat) | 48.6 | |
| 7 | MetaGPT (gpt-4o) | 48.6 | |
| 8 | AgentScope (gpt-4o) | 48.18 | |
| 9 | CAMEL (deepseek-chat) | 45.66 | |
| 10 | TimeClaw (deepseek-v4-pro) | 42.72 |
Interactive version: theaggregate.ai/benchmark?slug=temporalbench-t1 · How It Works · Data refreshed daily, snapshot 2026-10-20.