Back to Arena

PrivateGPT

by Zylon AI

System Card

OrganizationZylon AI
Released2023-05
Architecturevector-rag / FastAPI + LlamaIndex RAG, air-gapped
DetailsPrivacy-first RAG with FastAPI service and LlamaIndex-backed ingestion/retrieval pipelines. High-level APIs (ingest + chat) plus low-level primitives (embeddings, contextual retrieval); air-gapped capable.
Parameters—
Domainrag-retrieval
Open SourceYes
WebsiteVisit
CodeRepository
air-gappedfastapillamaindexprivacy

Capability Profile

Benchmark Scores

5 of 14 benchmarks
Long-Context Retrieval
2/5
RULER
69.334p
NIAH
no data
LooGLE
no data
LongBench
603p
∞Bench
no data
Multi-Turn Recall
0/2
LoCoMo
no data
MemoryBank
no data
Cross-Session Memory
0/1
LongMemEval
no data
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
65.324p
HotpotQA
60.815p
Agent Task Memory
0/1
AgentBench-Mem
no data
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
1/1
RAGAS
65.627p
Sources:PrivateGPT (zylon-ai/private-gpt); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)PrivateGPT (zylon-ai/private-gpt); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)PrivateGPT (zylon-ai/private-gpt); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)PrivateGPT (zylon-ai/private-gpt); evaluated on RAGAS: Automated Evaluation of Retrieval-Augmented Generation (Exploding Gradients, 2309)PrivateGPT (zylon-ai/private-gpt); evaluated on RULER: What's the Real Context Size of Your Long-Context Language Models (NVIDIA, 2404)