Back

Tether AI Research Group Unveils QVAC MedPsy Medical AI Model for Local Devices

According to Tether AI, QVAC MedPsy’s 1.7B and 4B models are optimized for local deployment, with quantized GGUF versions sized for privacy-sensitive hospital systems and mobile use cases.

USDT

Summary

Tether AI Research Group said it released QVAC MedPsy, a medical artificial intelligence model designed to run locally on smartphones, wearables, and other privacy-sensitive environments rather than relying on cloud-based processing. The company reported that the 1.7B version achieved a score of 62.62 across seven closed medical benchmarks, while the 4B version scored 70.54 and outperformed larger rival models. Tether AI also said recommended GGUF quantized versions are about 1.2GB for the 1.7B model and 2.6GB for the 4B model, positioning them for hospital and mobile deployments where local processing, compact model size, and data privacy are priorities.

Terms & Concepts
  • GGUF: A model file format commonly used for efficient local deployment and quantized inference of language models on consumer and edge devices.
  • Inference: The process of running an AI model on new inputs to generate outputs; local inference can reduce dependence on cloud infrastructure.
  • Benchmarks: Standardized tests used to measure a model’s performance across defined tasks or domains, such as medical evaluation sets.