Back to Arena

Open Interpreter

by OpenInterpreter

System Card

OrganizationOpenInterpreter
Released2023-07
Architectureagentic-workflow / Conversation-history memory with persistence hooks
DetailsLocal code-interpreter agent. Messages accumulated as model-executed code and outputs; conversation history can be reset or serialized (`interpreter.messages`) for save/restore; works with LM Studio, Ollama, Jan.
Parameters—
Domainagent-memoryepisodic-session
Open SourceYes
WebsiteVisit
CodeRepository
code-interpreteragpllocalconversation-memory

Capability Profile

Benchmark Scores

6 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
2/2
LoCoMo
75.349p
MemoryBank
74.540p
Cross-Session Memory
1/1
LongMemEval
76.353p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
72.961p
HotpotQA
72.759p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Open Interpreter (OpenInterpreter/open-interpreter); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Open Interpreter (OpenInterpreter/open-interpreter); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Open Interpreter (OpenInterpreter/open-interpreter); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Open Interpreter (OpenInterpreter/open-interpreter); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Open Interpreter (OpenInterpreter/open-interpreter); evaluated on MemoryBank: Enhancing LLMs with Long-Term Memory (Sun Yat-sen University, 2305)Open Interpreter (OpenInterpreter/open-interpreter); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)