Back to Arena

AutoGPT Platform

by Significant Gravitas

System Card

OrganizationSignificant Gravitas
Released2023-03
Architectureagentic-workflow / Continuous agent platform + Forge toolkit
DetailsLow-code frontend + backend runtime for continuous agents, with Forge SDK, agbenchmark, classic CLI. Agents hold state through scheduled executions; memory historically via pluggable vector stores.
Parameters—
Domainagent-memory
Open SourcePartial
WebsiteVisit
CodeRepository
autogptforgebenchmarkpolyform-shield

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
79.785p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
7970p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
73.766p
HotpotQA
79.595p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:AutoGPT Platform (Significant-Gravitas/AutoGPT); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)AutoGPT Platform (Significant-Gravitas/AutoGPT); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)AutoGPT Platform (Significant-Gravitas/AutoGPT); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)AutoGPT Platform (Significant-Gravitas/AutoGPT); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)AutoGPT Platform (Significant-Gravitas/AutoGPT); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)