PAST-Bench - Update: leaderboard
Metric: Capability task score (%; update families, where outdated state must be replaced; persistence on: the agent may reuse memory records, skills, profile state and session history produced by earlier episodes of the family; 26 scenarios and 204 episodes run as ordered sequences of fresh-session tasks; each evaluation episode scores completion (audit-data checks, or a MiniMax-M2.7 rubric judge for open-ended families), tool-error recovery and a safety gate; family-balanced means over three trials; Hermes agent framework). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GLM-5.1 | 90 |
| 2 | GPT-5.4 (Medium) | 80 |
| 3 | Kimi K2.6 (Thinking) | 78 |
| 4 | Claude Opus 4.6 (Medium) | 78 |
| 5 | Claude Sonnet 4.6 | 77 |
| 6 | DeepSeek V4 Pro (High) | 69 |
| 7 | MiniMax-M2.7 | 62 |
Interactive version: theaggregate.ai/benchmark?slug=past-bench-update · How It Works · Data refreshed daily, snapshot 2026-09-29.