Back to Arena

Qwen-Agent

by QwenLM (Alibaba)

System Card

OrganizationQwenLM (Alibaba)
Released2023-09
Architectureagentic-workflow / Qwen-native function calling + RAG memory
DetailsPowers Qwen Chat: LLMs, Tools, Agents with memory via function calling, MCP, code interpreter, and RAG. Docker code sandbox, parallel function calls, BrowserQwen app.
Parameters—
Domainagent-memoryrag-retrieval
Open SourceYes
WebsiteVisit
CodeRepository
qwenmcpragbrowserqwenalibaba

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
79.584p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
77.761p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
7250p
HotpotQA
70.849p
Agent Task Memory
1/1
AgentBench-Mem
7226p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
0/1
RAGAS
no data
Sources:Qwen-Agent (QwenLM/Qwen-Agent); evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Qwen-Agent (QwenLM/Qwen-Agent); evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)Qwen-Agent (QwenLM/Qwen-Agent); evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Qwen-Agent (QwenLM/Qwen-Agent); evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)Qwen-Agent (QwenLM/Qwen-Agent); evaluated on LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding (Tsinghua KEG, 2308)Qwen-Agent (QwenLM/Qwen-Agent); evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)