Bench'd
Claim
llamaindex-memory 0.0 on LoCoMollm-baseline 0.0 on LoCoMomem0-local 0.0 on LongMemEvalmem0-local 0.0 on LongMemEvalllamaindex-memory 0.0 on LongMemEvalllm-baseline 0.0 on LongMemEvallangchain-memory 0.0 on LongMemEvalcognee 0.0 on LongMemEval13 systems independently scored64 systems indexedllamaindex-memory 0.0 on LoCoMollm-baseline 0.0 on LoCoMomem0-local 0.0 on LongMemEvalmem0-local 0.0 on LongMemEvalllamaindex-memory 0.0 on LongMemEvalllm-baseline 0.0 on LongMemEvallangchain-memory 0.0 on LongMemEvalcognee 0.0 on LongMemEval13 systems independently scored64 systems indexed

MemPalace

Self-Reported
MemPalace LabsWebsiteGitHub(5.2k)DocsLast tested May 5, 2026

Spatial-metaphor memory system that organizes LLM memories into hierarchical rooms. Uses graph-based retrieval with episodic and semantic memory layers. Note: their claimed 96.6% score is retrieval recall (R@5), not end-to-end QA accuracy as measured by LongMemEval.

This system is indexed but hasn't been benchmarked yet.

Want to help? Run the Bench'd harness yourself Get started →

Command Palette

Search for a command to run...