MemoryBench (Meta)
In use
Models recall key facts from multi-turn conversations of 64K to 100K tokens that include distractor chats and turns that conflict with or update the facts. An LLM judge scores answers on weighted rubrics.
Meta in-house eval first reported in the Muse Wonder model card on 2026-10-12. Not the public MemoryBench by Ai et al. (ICML 2026).
Made by Meta (Meta Superintelligence Labs) Released Oct 2026 Finding facts in long inputs Tasks kept private Last in a launch post Oct 2026
Last model added Oct 2026
- Level
- Everyday Most adults, without training
- Scored
- Graded by a model A language model scores each response, usually against a rubric.
- Answers
- Open-ended Free-form text, documents, designs or media judged as a whole.
2
Launch posts list it, from 1 lab
2
System cards report it, from 1 lab
No scores here yet
No lab, leaderboard or Epoch AI run has reported a score for MemoryBench (Meta) yet.
One lab put it in a launch post in the last six months
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| 12 Oct 2026 |  Muse Wonder | 85.6 | MemoryBench |
| 12 Oct 2026 | Muse Wonder Full | 85.5 | MemoryBench |
2 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 12 Oct 2026 | Muse Wonder model card | 85.6 | MemoryBench, internal, Muse Wonder (max) (capabilities table image, page 3) | Capabilities |
| 12 Oct 2026 | Muse Wonder Full model card | 85.5 | MemoryBench, internal, Muse Wonder Full (max) (capabilities table image, page 3) | Capabilities |
More 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. Beam 1.0 66.9%
- 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 5 100%
- 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.
K-EXAONE 2.0 94.4% - 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%