AstroReason-Bench - Revisit Optimization - Revisit Gap: leaderboard
AstroReason-Bench (Wang et al., Fudan University, 2026) has LLM agents plan satellite operations in a physics-grounded simulator (orbit propagation, slewing, power and onboard storage) through tools and a Python API, all in the same Claude Code agent loop, over a four-day horizon. In the revisit task a constellation must keep observing monitoring targets while filling observation quotas for mapping targets and scheduling downlinks before storage runs out; this board scores the monitoring side as the average gap between consecutive observations of a target, in hours, over five generated cases. Lower is better.
Metric: Mean revisit gap over monitoring targets (hours; one run of five procedurally generated cases, Claude Code ReAct harness). Source: arxiv.org. Saturation forecast: Around February 2027. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 18.83 | #244 |
| 2 | Gemini 3 Flash | 24.96 | #159 |
| 3 | DeepSeek V3.1 NEX N1 | 26.78 | #858 |
| 4 | DeepSeek V3.2 | 29.89 | #403 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=astroreason-bench-revisit-optimization-revisit-gap · How It Works · Data refreshed daily, snapshot 2026-11-12.