MemoryArena: leaderboard
Metric: All-task average success rate (%): mean of the task success rates on Bundled Web Shopping, Group Travel Planning, Progressive Web Search, and math and physics formal reasoning; the model is the task agent and keeps its full interaction history in context (no external memory); printed as fractions and shown times 100; higher is better. Source: arxiv.org. Saturation forecast: Around July 2028. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 19 | #93 |
| 2 | Claude Sonnet 4.5 | 16 | #138 |
| 3 | GPT-4.1 Mini | 9 | #346 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=memoryarena · How It Works · Data refreshed daily, snapshot 2026-10-11.