TemporalBench - PSML T4: leaderboard
TemporalBench PSML process-systems slice for T4 event-conditioned forecasting questions.
Metric: Accuracy (%). Source: huggingface.co. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | TimeClaw (deepseek-v4-pro) | 50 | |
| 2 | TimeCopilot (deepseek-chat) | 49.12 | |
| 3 | Single LLM (deepseek-chat) | 36 | |
| 4 | Single LLM (gpt-4o) | 36 | |
| 5 | AgentScope (gpt-4o) | 35.33 | |
| 6 | AgentScope (deepseek-chat) | 31.33 | |
| 7 | MetaGPT (gpt-4o) | 31.33 | |
| 8 | CAMEL (deepseek-chat) | 30.67 | |
| 9 | CAMEL (gpt-4o) | 30.67 | |
| 10 | MetaGPT (deepseek-chat) | 30.67 |
Interactive version: theaggregate.ai/benchmark?slug=temporalbench-psml-t4 · How It Works · Data refreshed daily, snapshot 2026-10-20.