Skip to content
AI Model Radar

Fara1.5 27B

microsoft/Fara1.5-27B
Open Source · chat · open-weights
Open Alert me on changes
Context
262.1K
Max output
—
Weights
Open
API $/1M
—
Modalities
text · image
Released
17 Jul 2026
License: mit · microsoft/Fara1.5-27B
Download image Share on X Share on LinkedIn
AI summary
● machine-written

Fara1.5-27B: 27B computer use agent with 262K context window

Fara1.5-27B is a 27-billion parameter open-weights computer use agent model built on Qwen3.5, designed to complete complex browser-based tasks through an observe-think-act loop. The model achieves 72.3% on Online-Mind2Web benchmark and outperforms larger proprietary systems. It supports both text and image inputs with a 262K token context window and runs on a single A100/H100/B200 GPU.

What's new
  • Released as open-weights MIT-licensed model on Hugging Face
  • Achieves 72.3% on Online-Mind2Web, outperforming larger proprietary systems
  • Trained on FaraGen1.5 pipeline with synthetic domain data for gated/irreversible tasks
  • Operates without accessibility trees or separate parsing models
Best for
Complex multi-step browser automation tasksForm filling and data entry workflowsCross-site comparison shoppingTasks requiring reasoning about webpage state
Sources

Source: https://huggingface.co/microsoft/Fara1.5-27B