mlsanityData to 7 Oct 2026

GDPval: Evaluating AI Model Performance on Real-World Economically Valuable Tasks

arXiv 2510.04374 ↗First posted 5 Oct 2025cs.LG cs.AI cs.CYcited by 142

Not on the map. Why?

Cited by 68 in LLM agents and tool use, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

We introduce GDPval, a benchmark evaluating AI model capabilities on real-world economically valuable tasks. GDPval covers the majority of U.S. Bureau of Labor Statistics Work Activities for 44 occupations across the top 9 sectors contributing to U.S. GDP (Gross Domestic Product). Tasks are constructed from the representative work of industry professionals with an average of 14 years of experience. We find that frontier model performance on GDPval is improving roughly linearly over time, and that the current best frontier models are approaching industry experts in deliverable quality. We analyze the potential for frontier models, when paired with human oversight, to perform GDPval tasks cheaper and faster than unaided experts. We also demonstrate that increased reasoning effort, increased task context, and increased scaffolding improves model performance on GDPval. Finally, we open-source a gold subset of 220 tasks and provide a public automated grading service at evals.openai.com to facilitate future research in understanding real-world model capabilities.

Key facts. Cited by 142 papers in the map (data through 7 Oct 2026). Builds on 8 of its 28 references that are in the map. First posted on arXiv 5 Oct 2025.

Builds on

8 of 28 references.

References · 8

  1. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  2. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,748
  3. 2023AgentBench: Evaluating LLMs as Agentscited by 1,211
  4. 2024LLM Evaluators Recognize and Favor Their Own Generationscited by 651
  5. 2025Humanity's Last Examcited by 467
  6. 2023GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Modelscited by 174
  7. 2025SWE-Lancer: Can Frontier LLMs Earn 1 Million from Real-World Freelance Software Engineering?cited by 99
  8. 2024Clio: Privacy-Preserving Insights into Real-World AI Usecited by 66

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 8; 2026: 134.

142 papers.

Sort:

By area:

  1. 2026GLM-5: from Vibe Coding to Agentic Engineeringcited by 502
  2. 2026Kimi K2.5: Visual Agentic Intelligencecited by 449
  3. 2026Kimi K3: Open Frontier Intelligencecited by 99
  4. 2025NL2Repo-Bench: Towards Long-Horizon Repository Generation Evaluation of Coding Agentscited by 56
  5. 2026OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoningcited by 37
  6. 2026SWE-chat: Coding Agent Interactions From Real Users in the Wildcited by 36
  7. 2026Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoningcited by 32
  8. 2026OSWorld 2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Taskscited by 29
  9. 2026The 2025 AI Agent Index: Documenting Technical and Safety Features of Deployed Agentic AI Systemscited by 27
  10. 2026AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contextscited by 26
  11. 2026Gym-Anything: Turn any Software into an Agent Environmentcited by 24
  12. 2026Qwen-AgentWorld: Language World Models for General Agentscited by 22
38 more
  1. 2026Workspace-Bench 1.0: Benchmarking AI Agents on Workspace Tasks with Large-Scale File Dependenciescited by 20
  2. 2026SkillFlow:Benchmarking Lifelong Skill Discovery and Evolution for Autonomous Agentscited by 19
  3. 2025Remote Labor Index: Measuring AI Automation of Remote Workcited by 18
  4. 2026Reward Hacking in Rubric-Based Reinforcement Learningcited by 18
  5. 2025How Do AI Agents Do Human Work? Comparing AI and Human Workflows Across Diverse Occupationscited by 17
  6. 2026TurnOPD: Making On-Policy Distillation Turn-Aware for Efficient Long-Horizon Agent Trainingcited by 14
  7. 2025Intelligence per Watt: Measuring Intelligence Efficiency of Local AIcited by 13
  8. 2025AA-Omniscience: Evaluating Cross-Domain Knowledge Reliability in Large Language Modelscited by 12
  9. 2026SafePro: Evaluating the Safety of Professional-Level AI Agentscited by 10
  10. 2026How Well Does Agent Development Reflect Real-World Work?cited by 10
  11. 2026Automated Benchmark Auditing for AI Agents and Large Language Modelscited by 10
  12. 2026JobBench: Aligning Agent Work With Human Willcited by 10
  13. 2025Finch: Benchmarking Finance & Accounting across Spreadsheet-Centric Enterprise Workflowscited by 9
  14. 2026LLMs Corrupt Your Documents When You Delegatecited by 9
  15. 2026Nanbeige4.2-3B: Unlocking Agentic Capabilities in a Compact Modelcited by 9
  16. 2026AgentIF-OneDay: A Task-level Instruction-Following Benchmark for General AI Agents in Daily Scenarioscited by 8
  17. 2026EvoAgentBench: Benchmarking Agent Self-Evolution via Ability Transfercited by 8
  18. 2026Evo-Bench: Can Language Models Improve Agent Harness?cited by 8
  19. 2026EVMbench: Evaluating AI Agents on Smart Contract Securitycited by 7
  20. 2026AgenticVBench: Can AI Agents Complete Real-World Post-Production Tasks?cited by 7
  21. 2026CoffeeBench: Benchmarking Long-Horizon LLM Agents in Heterogeneous Multi-Agent Economiescited by 7
  22. 2026TUA-Bench: A Benchmark for General-Purpose Terminal-Use Agentscited by 7
  23. 2026Better Harnesses, Smaller Models: Building 90% Cheaper Agents via Automated Harness Adaptationcited by 7
  24. 2026WikiSkill: Compiling Agent Experience into Persistent Knowledge for Skill Evolutioncited by 7
  25. 2026BRIDGE: Predicting Human Task Completion Time From Model Performancecited by 6
  26. 2026OccuBench: Evaluating AI Agents on Real-World Professional Tasks via Language Environment Simulationcited by 6
  27. 2026CocoaBench: Evaluating Unified Digital Agents in the Wildcited by 6
  28. 2026Measuring Progress Toward AGI: A Cognitive Frameworkcited by 6
  29. 2026BigFinanceBench: A Workflow-Grounded Benchmark for Financial-Research Agentscited by 6
  30. 2026Monitorability as a Free Gift: How RLVR Spontaneously Aligns Reasoningcited by 4
  31. 2026ServImage: An Image Generation and Editing Benchmark from Real-world Commercial Imaging Servicescited by 4
  32. 2026MBABench: Evaluating LLM Agents on End-to-End Spreadsheet Tasks in Financecited by 4
  33. 2026Workflow-GYM: Towards Long-Horizon Evaluation of Computer-use Agentic tasks in Real-World Professional Fieldscited by 4
  34. 2026CEO-Bench: Can Agents Play the Long Game?cited by 4
  35. 2026Sakana Fugu Technical Reportcited by 4
  36. 2026Hidden Decoding at Scale: Latent Computation Scaling for Large Language Modelscited by 4
  37. 2026NexForge: Scaling Agent Capabilities through Requirement-Driven Task Synthesis for LLMscited by 4
  38. 2026Basic Legibility Protocols Improve Trusted Monitoringcited by 3

50 most cited of 142 All 142 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026The Capability Frontier: Benchmarks Miss 82% of Model Performance
  2. 2025General Scales Unlock AI Evaluation with Explanatory and Predictive Power
  3. 2025Scaling Laws for Economic Productivity: Experimental Evidence in LLM-Assisted Consulting, Data Analyst, and Management Tasks
  4. 2024RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts
  5. 2026FrontierScience: Evaluating AI's Ability to Perform Expert-Level Scientific Tasks
  6. 2025GrandJury: A Collaborative Machine Learning Model Evaluation Protocol for Dynamic Quality Rubrics
  7. 2025Measurement to Meaning: A Validity-Centered Framework for AI Evaluation
  8. 2025Can AI Master Econometrics? Evidence from Econometrics AI Agent on Expert-Level Tasks