Back to Arena

MongoDB Atlas Vector

by MongoDB Inc.

System Card

OrganizationMongoDB Inc.
Released2023-06
Architecturevector-rag / document-db-native-vector
DetailsAtlas Vector Search adds HNSW-indexed semantic search directly inside MongoDB Atlas documents, enabling queries that combine vector similarity with aggregation pipeline filters, geo queries, and relational lookups without data export. Named the most-loved vector database in Retool's 2024 State of AI Report. Acquired Voyage AI for $220M to pair best-in-class embeddings with the store.
Parameters—
Domainrag-retrievalagent-memory
Open SourceNo
WebsiteVisit
document-dbatlashnswaggregation-pipelinemanaged

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
55.65p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
575p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
60.13p
HotpotQA
63.525p
Agent Task Memory
1/1
AgentBench-Mem
63.812p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:MongoDB Atlas Vector vendor documentation; evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)MongoDB Atlas Vector vendor documentation; evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)MongoDB Atlas Vector vendor documentation; evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)MongoDB Atlas Vector vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)MongoDB Atlas Vector vendor documentation; evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)MongoDB Atlas Vector vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)