@ms_aifrontiers: Fara1.5 is here! The tech report just landed on arXiv. New SOTA for computer use agents of its size, and it competes wi…

X AI KOLs Following Models

Summary

Fara1.5 is a family of native computer use agents trained using the FaraGen1.5 scalable data pipeline. The models achieve new state-of-the-art results on browser-use benchmarks, competing with much larger frontier models.

Fara1.5 is here! The tech report just landed on arXiv. New SOTA for computer use agents of its size, and it competes with much larger frontier models. Paper: https://t.co/BkhgwNuxiq
Original Article
View Cached Full Text

Cached at: 06/25/26, 09:16 AM

Fara1.5 is here!

The tech report just landed on arXiv. New SOTA for computer use agents of its size, and it competes with much larger frontier models.

Paper: https://t.co/BkhgwNuxiq


Fara-1.5: Scalable Learning Environments for Computer Use Agents

Source: https://arxiv.org/abs/2606.20785 Authors:Ahmed Awadallah,Sahil Gupta,Yash Lara,Yadong Lu,Hussein Mozannar,Akshay Nambi,Zach Nussbaum,Yash Pandya,Aravind Rajeswaran,Corby Rosset,Alexey Taymanov,Luiz do Valle,Vibhav Vineet,Spencer Whitehead,Andrew Zhao

View PDF

Abstract:Collecting computer use data from human demonstrations is expensive and slow, motivating the need for scalable generation strategies. This requires two key ingredients: environments in which agents can act and verifiers that can judge whether their demonstrations succeeded. We introduce FaraGen1.5, a scalable data pipeline for computer use agents composed of three modular components: environments, solvers, and verifiers. FaraGen1.5 uses both live websites and synthetic environments that faithfully simulate domains gated by authentication or that require irreversible actions. It employs a solver harness that can be powered by multiple models, including strong frontier models such as GPT-5.4, and also incorporates a user simulator to enable multi-turn rollouts. Finally, FaraGen1.5 scores the resulting trajectories with three complementary verifiers covering task correctness, efficiency, and critical-point adherence. Using data produced by this pipeline, we train Fara1.5, a family of native computer use agents (CUAs) at three scales built on Qwen3.5 (4B, 9B, and 27B). To train these models, we employ a supervised finetuning (SFT) recipe that carefully balances data from FaraGen1.5 for broad coverage, specific high-value tasks, and target model deficiencies in an iterative approach. Each model sets a new state of the art for its size class on browser-use benchmarks: Fara1.5-9B reaches 63.4% on Online-Mind2Web and 86.6% on WebVoyager, while Fara1.5-27B achieves 72.3% on Online-Mind2Web, which is competitive with much larger proprietary systems.

Submission history

From: Aravind Rajeswaran [view email] **[v1]**Thu, 18 Jun 2026 17:53:03 UTC (12,657 KB)

Similar Articles

Fara-7B: An Efficient Agentic Model for Computer Use

Papers with Code Trending

Introduces FaraGen, a synthetic data generation system for computer use agents, and Fara-7B, a small but efficient model that outperforms larger counterparts on web task benchmarks. The model is released open-weight on Microsoft Foundry and HuggingFace.

microsoft/Fara-7B

Hugging Face Models Trending

Microsoft released Fara-7B, an efficient 7 billion parameter agentic small language model (SLM) for computer use tasks, achieving state-of-the-art performance within its size class and competitive with larger systems.