Cognitapp/Cognitapp-Med-Nano-v1

Hugging Face
TEXT GENERATIONConcurrency Cost:1Model Size:0.5BQuant:BF16Ctx Length:32kPublished:Mar 8, 2026License:apache-2.0Architecture:Transformer0.0K Open Weights Warm

Cognitapp/Cognitapp-Med-Nano-v1 is a specialized 0.5 billion parameter medical large language model developed by Cognitapp Labs. Fine-tuned from the Qwen2.5-0.5B architecture, it excels at ICD-10-CM Medical Billing and Clinical Extraction. This model is optimized for 100% offline use on mobile and desktop devices via MLX or llama.cpp, prioritizing alphanumeric code accuracy through prompt-masking. It is designed as a supportive tool for medical professionals and billers, with global and regional clinical awareness.

Loading preview...

Cognitapp-Med-Nano-v1: Specialized Medical LLM

Cognitapp-Med-Nano-v1 is a lightweight, specialized medical large language model (LLM) developed by Cognitapp Labs. Built upon the Qwen2.5-0.5B architecture, this 0.5 billion parameter model is specifically fine-tuned for ICD-10-CM Medical Billing and Clinical Extraction.

Key Capabilities & Features

  • Medical Specialization: Optimized for accurate ICD-10-CM code extraction and clinical information processing.
  • Efficiency & Offline Use: With only 0.5 billion parameters, it is designed for efficient, 100% offline operation on mobile and desktop devices using frameworks like MLX or llama.cpp.
  • Precision: Utilizes prompt-masking during training to enhance alphanumeric code accuracy, reducing conversational filler.
  • Global & Regional Awareness: Trained to incorporate both international and regional clinical standards.
  • Training Data: Fine-tuned on over 1,200 diverse clinical scenarios, covering pediatrics, geriatrics, and infectious diseases.

Intended Use

This model serves as a supportive tool for medical professionals and billers, aiding in the extraction of medical codes and clinical data. It is explicitly not intended as a diagnostic tool, and all outputs require verification by a licensed healthcare professional.