Back to Arena

Onyx

by onyx-dot-app

System Card

Organizationonyx-dot-app
Released2023-01
Architecturehybrid / Enterprise search + agentic RAG
DetailsEnterprise search/chat layer supporting agentic RAG, deep research, custom agents, web search, code execution, voice, with enterprise features (SSO/RBAC/audit). Chat sessions carry persistent memory.
Parameters—
Domainrag-retrievalagent-memory
Open SourceYes
WebsiteVisit
CodeRepository
enterprise-searchdeep-researchssorbac

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
1/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
603p
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
7658p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
79.677p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
74.575p
HotpotQA
7994p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Onyx (onyx-dot-app/onyx); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Onyx (onyx-dot-app/onyx); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)Onyx (onyx-dot-app/onyx); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Onyx (onyx-dot-app/onyx); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Onyx (onyx-dot-app/onyx); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)Onyx (onyx-dot-app/onyx); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)