Letta Memory Benchmark - Archival Memory Read: leaderboard
The Letta agent must search its archival memory to answer a question, over 100 items repeated 3 times; the score is the share answered correctly. It is the retrieval part of the Letta Memory Benchmark.
Metric: Score (%) on 100 archival-memory read 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 3.5 Haiku (20241022) | 100 | #1601 | |
| 2 | GPT-4.1 Mini | 99.67 | #781 | |
| 3 | Gemini 2.5 Pro (Preview 05-06) | 98.67 | #279 | |
| 4 | Claude Sonnet 4 | Thinking | 98.33 | #406 |
| 5 | GPT-4.1 Nano | 94.33 | #2035 | |
| 6 | GPT-4o (2024-08-06) | 92.33 | #736 | |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 92.33 | #423 | |
| 8 | Gemini 2.5 Flash (Preview 04-17) | 89 | #458 | |
| 9 | GPT-4o Mini (2024-07-18) | 82.67 | #1511 | |
| 10 | Claude Sonnet 4 (20250514) | 81.33 | #388 | |
| 11 | Llama 4 Scout Instruct | 64.33 | #1579 | |
| 12 | Llama 4 Maverick Instruct FP8 | 49.33 | #1010 | |
| 13 | Claude 3.7 Sonnet (20250219) | 47 | #414 | |
| 14 | Llama 3.2 Instruct Turbo (3B) | 12.67 | ||
| 15 | Llama 3.3 70B Instruct | 8.33 | #1379 |
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-archival-memory-read · How It Works · Data refreshed daily, snapshot 2026-11-16.