Benchmark Almanac

Agent Memory Leaderboard

MemoraX (memory system) 58.0 tops its leaderboard
Not in model cards

Memory systems for agents are evaluated on storing, retrieving and using long-term memory across textual, coding and multimodal tracks in one reproducible framework.

Runs a 2026 challenge in cycles. Ranks memory systems, not bare models. About 1,100 GitHub stars and 68 forks.

Made by
AML-memory
Released
July 2026
Kind
Learning from the context
Level
Everyday
Scored
Checked automatically
Answers
Agent task
Tasks
All public
15
Entries on its own leaderboard, 0 models we track

MemoraX (memory system) tops the leaderboard with 58.0

Best score per entry from the benchmark's own leaderboard, read on November 12th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

15 entries. Scroll sideways for the rest.

  1. 58.0 MemoraX (memory system)
  2. 45.9 MemOS (memory system)
  3. 44.2 NTES-MEMORY-SMART (memory system)
  4. 43.6 AML-Eval-FLASH (memory system)
  5. 42.6 Cognee (memory system)
  6. 42.1 NTES-MEMORY-MQ-LAYER (memory system)
  7. 41.5 TencentDB (memory system)
  8. 41.4 Mem0 (memory system)
  9. 39.5 MemPalace (memory system)
  10. 38.5 Vectorize Hindsight Cloud (memory system)
  11. 37.6 memory-dense (memory system)
  12. 35.6 星际罗盘 (memory system)
  13. 34.2 memory-8000 (memory system)
  14. 29.6 SuperMemory (memory system)
  15. 26.3 NTES-MEMORY-TDAM (memory system)
  1. 1 MemoraX (memory system) v0.5 58.0
  2. 2 MemOS (memory system) v2.0.1 45.9
  3. 3 NTES-MEMORY-SMART (memory system) AML-Eval-SMART 44.2
  4. 4 AML-Eval-FLASH (memory system) FLASH 43.6
  5. 5 Cognee (memory system) v1.4.0 42.6
  6. 6 NTES-MEMORY-MQ-LAYER (memory system) v1.0 42.1
  7. 7 TencentDB (memory system) v2.0.0 41.5
  8. 8 Mem0 (memory system) v3 41.4
  9. 9 MemPalace (memory system) v3.5.0 39.5
  10. 10 Vectorize Hindsight Cloud (memory system) v1.0.1 38.5
  11. 11 memory-dense (memory system) v1.1 37.6
  12. 12 星际罗盘 (memory system) v1.0 35.6
  13. 13 memory-8000 (memory system) v1.0 34.2
  14. 14 SuperMemory (memory system) v3write v4search 29.6
  15. 15 NTES-MEMORY-TDAM (memory system) TDAM V1 26.3

Source: Official Agent Memory Leaderboard leaderboard

#EntryScoreSettingDateSource
1 MemoraX (memory system) 58.0 v0.5 Official leaderboard
2 MemOS (memory system) 45.9 v2.0.1 Official leaderboard
3 NTES-MEMORY-SMART (memory system) 44.2 AML-Eval-SMART Official leaderboard
4 AML-Eval-FLASH (memory system) 43.6 FLASH Official leaderboard
5 Cognee (memory system) 42.6 v1.4.0 Official leaderboard
6 NTES-MEMORY-MQ-LAYER (memory system) 42.1 v1.0 Official leaderboard
7 TencentDB (memory system) 41.5 v2.0.0 Official leaderboard
8 Mem0 (memory system) 41.4 v3 Official leaderboard
9 MemPalace (memory system) 39.5 v3.5.0 Official leaderboard
10 Vectorize Hindsight Cloud (memory system) 38.5 v1.0.1 Official leaderboard
11 memory-dense (memory system) 37.6 v1.1 Official leaderboard
12 星际罗盘 (memory system) 35.6 v1.0 Official leaderboard
13 memory-8000 (memory system) 34.2 v1.0 Official leaderboard
14 SuperMemory (memory system) 29.6 v3write v4search Official leaderboard
15 NTES-MEMORY-TDAM (memory system) 26.3 TDAM V1 Official leaderboard

No lab model card reports it yet

We read the headline results of 522 launch posts from 52 labs, and none of them list this benchmark.

More in Long context

All 79 in Long context
  • MRCR v2 In use Multi-round coreference: retrieve the n-th of several similar requests hidden in a long synthetic conversation, up to 1M tokens. Muse Spark 1.3 98.5%
  • LongBench v2 In use 503 multiple-choice questions over 8K-2M-word contexts (QA, dialogue, code repos, structured data). Accuracy. Inkling 2 Small 70.6%
  • GraphWalks In use A graph is given as an edge list in long context. The model answers BFS or parent queries, scored by F1. Claude Fable 5 99.96%
  • OpenAI MRCR v2 In use Models read a long synthetic chat that hides 2, 4 or 8 identical writing requests and must reproduce the response to the i-th one. Scored by mean string match ratio per context bin. Gemini 3.7 Flash 89%
  • CL-bench In use Model must learn new knowledge from a supplied long context (rules, procedures, data) and solve tasks. Strict LLM-verified rubrics. GPT-5.4 27.9%
  • AA-LCR 1.1 In use Models answer 100 questions that need reasoning across sets of real documents of about 100K tokens. An LLM equality checker with a fixed system prompt compares each answer with the key. Step 5 Preview 88.3%