CrowtherLabs/atom-proton-1.0

VISIONPricing:Input $1.6 / Cached $0.15 / Output $12Concurrent Unit Cost:2Model Size:27BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 16, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

CrowtherLabs/atom-proton-1.0 is a 27 billion parameter vision-language model developed by CrowtherLabs, based on Qwen3.8-27B-FP8. It is designed to read text and images, perform complex reasoning, and draft professional documents. This model is optimized for enterprise workflows, excelling at tasks like extracting document structure, reviewing written work, and carrying out ordinary reasoning in long-horizon search and reasoning tasks.

Loading preview...

Atom Proton 1.0 Overview

Atom Proton 1.0 is a 27 billion parameter vision-language model developed by CrowtherLabs, designed for enterprise applications. It is built upon the Qwen3.8-27B-FP8 base model and has undergone proprietary post-training to enhance its capabilities in long-horizon tasks. This model is adept at processing both text and images, performing complex reasoning, and generating professional-grade written content.

Key Capabilities

  • Vision-Language Understanding: Processes and interprets information from both textual and image inputs.
  • Advanced Reasoning: Engineered to reason before generating responses, suitable for complex enterprise workflows.
  • Professional Drafting: Capable of drafting and reviewing written work, maintaining a professional register.
  • Document Structure Extraction: Excels at extracting structured information from various documents.
  • Optimized for Enterprise: Specifically designed for tasks like those found in BrowseComp Plus submissions, focusing on long-horizon search and reasoning.

Good for

  • Enterprise Workflows: Ideal for applications requiring sophisticated document processing, content generation, and reasoning within corporate environments.
  • Complex Information Extraction: Suitable for tasks that involve extracting detailed structures and insights from large volumes of text and images.
  • Automated Content Generation: Useful for drafting reports, summaries, and other professional documents where a specific tone and accuracy are required.
  • Research and Development: As a research artifact, it serves as a strong foundation for further development in vision-language understanding and reasoning systems.

Finetunes
1 models