Letta Memory Benchmark - Core Memory Update: leaderboard
The Letta agent must update a fact held in its core memory block, over 100 items repeated 3 times; the score is the share done correctly. It differs from the core memory read task in requiring an edit through the memory tool rather than an answer from context.
Metric: Score (%) on 100 core-memory update items, 3 repeats, Letta agent framework. Source: raw.githubusercontent.com. Saturation forecast: Estimated already saturated. 18 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Sonnet 4 (20250514) | 83.67 | #388 | |
| 2 | Claude 3.7 Sonnet (20250219) | 76.33 | #414 | |
| 3 | GPT-4o (2024-08-06) | 73 | #736 | |
| 4 | Claude Sonnet 4 | Thinking | 72.33 | #406 |
| 5 | GPT-4o Mini (2024-07-18) | 70.67 | #1511 | |
| 6 | O4 Mini | 66.33 | #356 | |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 61.33 | #423 | |
| 8 | GPT-4.1 Mini | 51.33 | #781 | |
| 9 | Gemini 2.5 Flash (Preview 04-17) | 49.33 | #458 | |
| 10 | Llama 4 Maverick Instruct FP8 | 46.67 | #1010 | |
| 11 | Gemini 2.5 Pro (Preview 05-06) | 46.67 | #279 | |
| 12 | Llama 3.3 70B Instruct | 44.33 | #1379 | |
| 13 | Claude 3.5 Haiku (20241022) | 28.33 | #1601 | |
| 14 | O3 Mini | 16.67 | #544 | |
| 15 | Llama 4 Scout Instruct | 13.67 | #1579 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=letta-memory-benchmark-core-memory-update · How It Works · Data refreshed daily, snapshot 2026-11-16.