![[2601.14242] APEX-Agents](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2601.14242] APEX-Agents
Summary
The APEX-Agents paper introduces a benchmark for evaluating AI agents' ability to perform complex tasks across various applications, showcasing the performance of multiple AI models.
Why It Matters
As AI agents become increasingly integrated into professional environments, the APEX-Agents benchmark provides a necessary framework for assessing their effectiveness in executing real-world tasks. This research is crucial for developers and organizations aiming to improve AI productivity and reliability.
Key Takeaways
- APEX-Agents benchmark evaluates AI agents on long-horizon tasks.
- Gemini 3 Flash achieved the highest score of 24.0% among tested agents.
- The benchmark is open-sourced, promoting transparency and collaboration.
- It includes realistic work environments to test agent capabilities.
- The paper contributes to the ongoing development of AI productivity tools.
Computer Science > Computation and Language arXiv:2601.14242 (cs) [Submitted on 20 Jan 2026 (v1), last revised 23 Feb 2026 (this version, v3)] Title:APEX-Agents Authors:Bertie Vidgen, Austin Mann, Abby Fennelly, John Wright Stanly, Lucas Rothman, Marco Burstein, Julien Benchek, David Ostrofsky, Anirudh Ravichandran, Debnil Sur, Neel Venugopal, Alannah Hsia, Isaac Robinson, Calix Huang, Olivia Varones, Daniyal Khan, Michael Haines, Austin Bridges, Jesse Boyle, Koby Twist, Zach Richards, Chirag Mahapatra, Brendan Foody, Osvald Nitski View a PDF of the paper titled APEX-Agents, by Bertie Vidgen and 23 other authors View PDF Abstract:We introduce the AI Productivity Index for Agents (APEX-Agents), a benchmark for assessing whether AI agents can execute long-horizon, cross-application tasks created by investment banking analysts, management consultants, and corporate lawyers. APEX-Agents requires agents to navigate realistic work environments with files and tools. We test eight agents for the leaderboard using Pass@1. Gemini 3 Flash (Thinking=High) achieves the highest score of 24.0%, followed by GPT-5.2 (Thinking=High), Claude Opus 4.5 (Thinking=High), and Gemini 3 Pro (Thinking=High). We open source the APEX-Agents benchmark (n=480) with all prompts, rubrics, gold outputs, files, and metadata. We also open source Archipelago, our infrastructure for agent execution and evaluation. Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs...
