Back to Arena

SuperAGI

by TransformerOptimus

System Card

OrganizationTransformerOptimus
Released2023-05
Architectureagentic-workflow / Autonomous agent platform with vector memory
DetailsEnd-to-end framework for autonomous agents with multi-vector-DB support (Pinecone, Qdrant, Weaviate), GUI action console, tool integrations, and vector-memory-based learning across runs.
Parameters—
Domainagent-memory
Open SourceYes
WebsiteVisit
CodeRepository
autonomousmulti-vectordbgui-consoledeprecated-tag

Capability Profile

Benchmark Scores

5 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
76.261p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
85.195p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
7579p
HotpotQA
77.385p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:SuperAGI (TransformerOptimus/SuperAGI); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)SuperAGI (TransformerOptimus/SuperAGI); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)SuperAGI (TransformerOptimus/SuperAGI); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)SuperAGI (TransformerOptimus/SuperAGI); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)SuperAGI (TransformerOptimus/SuperAGI); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)