Continual Learning Bench: leaderboard
Agent benchmark for learning across sequences of related tasks, measuring whether systems improve from prior attempts instead of solving each task independently.
Metric: Agg. Reward (%). Source: continual-learning-bench.com. Saturation forecast: Around 2030. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.6 | 19.6 | #90 |
| 2 | GPT-5.4 | 18.9 | #79 |
| 3 | Claude Opus 4.7 | 18.32 | #49 |
| 4 | Gemini 3 Flash | 9.22 | #99 |
| 5 | Gemini 3.1 Pro (Preview) | 0.33 | #58 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=continual-learning-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.