LifeSearch: Benchmarking Deep Research Agents for Real-World Life and Productivity

Hugging Face Daily Papers Papers
benchmark deep-research-agents evaluation llm-judges multilingual productivity

Summary

LifeSearch introduces a benchmark of 203 Chinese and English tasks for evaluating deep research agents in real-world life and productivity settings, with over 4,000 expert-written rubrics and experiments on 12 advanced agents, finding that even the best frontier agent (GPT-5.6 Sol) scores only 63.4%.

Deep research (DR) agents address open-ended information needs through multi-step reasoning, long-horizon web search, and evidence synthesis. However, existing benchmarks primarily focus on academic research or synthetic tasks, leaving DR agents' capabilities in real-world life and professional productivity settings insufficiently evaluated. To bridge this gap, we introduce LifeSearch, a challenging benchmark comprising 203 Chinese and English tasks derived from real-user query seeds across 7 concrete life and productivity domains. With more than 4,500 hours of human annotation, we construct over 4,000 expert-written, fine-grained rubrics, covering both report-content criteria for coverage and accuracy and concrete criteria specifying values, units, dates, and URLs. We conduct extensive experiments on 12 advanced DR agents under a unified evaluation harness, systematically analyzing their performance patterns and failure modes. We find that even the best-performing frontier agent, GPT-5.6 Sol, achieves a score of only 63.4%, leaving substantial room for improvement on realistic life and professional DR tasks. We select the grading configuration by comparing candidate LLM judges against human rubric labels, and a human ranking study shows strong alignment between aggregate benchmark scores and human judgments (Spearman's ρ= 0.89, Kendall's τ= 0.81). LifeSearch provides a rigorous evaluation framework and foundation for advancing DR agents toward real-world life and productivity.
Original Article
View Cached Full Text

Cached at: 10/14/26, 10:51 AM

Paper page - LifeSearch: Benchmarking Deep Research Agents for Real-World Life and Productivity

Source: https://huggingface.co/papers/2610.18092

Abstract

Deepresearch(DR)agentsaddressopen-endedinformationneedsthroughmulti-stepreasoning,long-horizonwebsearch,andevidencesynthesis.However,existingbenchmarksprimarilyfocusonacademicresearchorsynthetictasks,leavingDRagents’capabilitiesinreal-worldlifeandprofessionalproductivitysettingsinsufficientlyevaluated.Tobridgethisgap,weintroduceLifeSearch,achallengingbenchmarkcomprising203ChineseandEnglishtasksderivedfromreal-userqueryseedsacross7concretelifeandproductivitydomains.Withmorethan4,500hoursofhumanannotation,weconstructover4,000expert-written,fine-grainedrubrics,coveringbothreport-contentcriteriaforcoverageandaccuracyandconcretecriteriaspecifyingvalues,units,dates,andURLs.Weconductextensiveexperimentson12advancedDRagentsunderaunifiedevaluationharness,systematicallyanalyzingtheirperformancepatternsandfailuremodes.Wefindthateventhebest-performingfrontieragent,GPT-5.6Sol,achievesascoreofonly63.4%,leavingsubstantialroomforimprovementonrealisticlifeandprofessionalDRtasks.WeselectthegradingconfigurationbycomparingcandidateLLMjudgesagainsthumanrubriclabels,andahumanrankingstudyshowsstrongalignmentbetweenaggregatebenchmarkscoresandhumanjudgments(Spearman’sρ=0.89,Kendall’sτ=0.81).LifeSearchprovidesarigorousevaluationframeworkandfoundationforadvancingDRagentstowardreal-worldlifeandproductivity.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2610\.18092

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/2610.18092 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2610.18092 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2610.18092 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

LifeSide: Benchmarking Agents as Lifelong Digital Companions

arXiv cs.CL

LifeSide is a new benchmark for evaluating AI agents as lifelong digital companions, testing memory tracking, user understanding, privacy control, and emotional companionship across 2,000 personas and 111K tasks in multi-session settings. Results show that even top models fail to sustain accurate user understanding and genuine companionship over long horizons.

Introducing LifeSciBench

OpenAI Blog

OpenAI introduces LifeSciBench, a benchmark of 750 expert-authored tasks to evaluate AI systems on realistic life science research workflows, including evidence handling, analysis, and scientific reasoning.