Quaid Evals

quaid-app.github.io/quaid-evals · 4 runs

Latest: 0.9.8 (2026-04-27)

DAB Score
No runs yet
P@5 (gbrain-evals)
38.48%
GBrain ref: 49.1%
R@5 (gbrain-evals)
173.4%
GBrain ref: 97.9%

Score History

Competitive Comparison

System DAB P@5 R@5 LoCoMo BEAM 1M BEAM 10M Notes
Quaid (latest) ours 38.48% 173.4% PARA-native, OpenClaw MCP, self-hosted
GBrain 49.1% 97.9% Graph+vector+FTS, 17,888 pages, 2026-04-27
Mem0 v3 91.6% 64.1% 48.6% ~7K tokens/retrieval, conversational memory, 2026-04-01

Reference scores from published benchmarks. Different benchmarks test different capabilities — see methodology.

All Runs

Version Date DAB P@5 R@5
0.9.8 2026-04-27 38.48% 173.4%
0.9.10 2026-04-27 17.44% 78.17%
0.9.8 2026-04-27
0.9.10 2026-04-27