# OmniMemEval

> MemTensor's evaluation framework for memory backends (LoCoMo, LongMemEval, BEAM, PersonaMem v2, HaluMem) and memory-equipped agents on AgentBench domains.

- id: `omnimemeval`
- category: Evals & Observability (`evals-observability`)
- url: https://github.com/MemTensor/OmniMemEval
- repo: https://github.com/MemTensor/OmniMemEval
- status: active
- pricing: open-source
- license: Apache-2.0
- tags: evals, memory, benchmarks, agents, python
- related: MemOS (https://indexagentica.com/entries/memos/); Mem0 (https://indexagentica.com/entries/mem0/); AgentBench (https://indexagentica.com/entries/agentbench/); OpenClaw (https://indexagentica.com/entries/openclaw/)
- sources: https://github.com/MemTensor/OmniMemEval
- added: 2026-10-27
- updated: 2026-10-27
- last verified: 2026-10-27
- featured in: Memory servers for coding agents compared: m3 Memory vs agentmemory vs Basic Memory vs MCP Memory Service (https://indexagentica.com/compare/coding-agent-memory-servers/)
- maintainer: MemTensor
- submitted by: agentica-curator
- page: https://indexagentica.com/entries/omnimemeval/
- json: https://indexagentica.com/api/entries/omnimemeval.json

Two tracks: **User Memory Evaluation** drives any memory backend through `add()` and `search()` and scores answers with configurable answer/eval LLMs; **Agent Memory Evaluation** runs agent runtimes with memory plugins (OpenClaw) on AgentBench-style domains (reasoning, information retrieval, knowledge work, code, software engineering). Ships data-preparation scripts, env templates per backend (e.g. MemOS) and published result snapshots. Third-party benchmark data keeps its upstream licenses. From the makers of MemOS.