LongCat-DeepResearch Technical Report
Summary
LongCat-DeepResearch is a deep research system that combines an enhanced LongCat model with a multi-agent workflow for producing comprehensive, evidence-grounded reports, achieving competitive scores on benchmarks like DeepResearchBench and ResearchRubrics.
View Cached Full Text
Cached at: 09/30/26, 04:14 AM
Paper page - LongCat-DeepResearch Technical Report
Source: https://huggingface.co/papers/2609.36071 Authors:
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
,
Abstract
WepresentLongCat-DeepResearch,adeepresearchsystemthatcombinesanenhancedLongCatmodelwithamulti-agentworkflowforproducingcomprehensive,evidence-groundedreports.Theworkflowseparatesglobalplanningfromdetailedinvestigationandcoordinatesrevisionatthesectionlevel.Multipleplanningagentsfirstexploreexternalsourcesandrefineanactionableresearchplan,termedResearchSpec.Researchagentstheninvestigateanddrafttheirassignedsectionsinparallel,gatheringadditionalevidenceinseparatecontextsastheiranalysesdevelop.Oncethesectionsareassembled,globalreviewguidestargetedlocalrevisions,reducingrelianceonrepeatedfull-reportrewriting.Thisworkflowalsosupportstheconstructionofresearchtasksandtrajectoriesforthemid-trainingandpost-trainingofLongCat’sgeneral-purposemodels.LongCat-DeepResearchachieves55.25onDeepResearchBench,51.35onDeepResearchBenchII,and79.83onResearchRubrics.Onanin-housebenchmark,itscores76.04,rankingsecondamongfourcomparedsystems.Development-setanalysesshowbenefitsfromcombiningplanningperspectives,whilefurtherplanningrefinementhasmixedeffects.Additionaleditingimprovesaverageautomaticreadabilitypreferenceacrosstwobenchmarks,withdifferenttrendsoneach.
View arXiv pageView PDFAdd to collection
Get this paper in your agent:
hf papers read 2609\.36071
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2609.36071 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2609.36071 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2609.36071 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
Mind DeepResearch Technical Report
MindDR is a multi-agent deep research framework using a three-agent architecture (Planning, DeepSearch, Report) and a four-stage training pipeline, achieving competitive performance with ~30B-parameter models on multiple benchmarks. Developed by Li Auto and deployed as an online product, it also introduces MindDR Bench, a 500-query Chinese benchmark for evaluating deep research capabilities.
langchain-ai/open_deep_research
Open Deep Research is a configurable, fully open source deep research agent by LangChain that supports multiple model providers, search tools, and MCP servers, achieving competitive performance on the Deep Research Bench leaderboard.
DuMate-DeepResearch: An Auditable Multi-Agent System with Recursive Search and Rubric-Grounded Reasoning
This technical report introduces DuMate-DeepResearch, a multi-agent framework for deep research tasks that decouples the agent core from a tool ecosystem, and incorporates graph-based dynamic planning, recursive two-level execution, and rubric-based test-time optimization. The system achieves state-of-the-art results on two deep research benchmarks, demonstrating the value of auditable agent infrastructure.
@DanKornas: Complex research agents get messy fast: planning, search, RAG, code execution, feedback, and final reports all need to …
DeepResearch is an open-source multi-agent research tool built with Spring AI Alibaba that converts queries into structured reports using dynamic planning, multi-agent roles, hybrid RAG, and Docker-based execution.
S1-DeepResearch: Beyond Search, Toward Real-World Long-Horizon Research Agents
This paper introduces S1-DeepResearch-32B, an open-source model and 15K trajectory dataset for deep research agents, achieving state-of-the-art performance across 20 benchmarks by jointly modeling information acquisition, knowledge synthesis, and planning.