Ling-3.0-flash-Sante

Researchers and developers can use it for healthcare question answering, literature and evidence synthesis, long-document analysis, and tool-assisted research with appropriate human review.

Availability

Model weights available on Hugging Face and ModelScope

Pricing model

Not established

Deployment

Quickstarts provided for self-hosting with SGLang and vLLM

Privacy

Users are responsible for protecting personal health information and complying with applicable privacy requirements

Licence

MIT

• The model is designed for medical reasoning, healthcare research, deep research, and tool-assisted workflows. ^[1]

  With **124B total parameters** and **5.1B active parameters per token**, Sante is designed for **medical reasoning, healthcare research, deep research, and tool-assisted workflows** while retaining the general capabilities of Ling-3.0-flash.

• Training extends to 256K tokens. ^[1]

  + **Built for long contexts:** Training extends to 256K tokens, supporting long-document analysis, evidence synthesis, and multi-step research.

• Model weights are available on Hugging Face and ModelScope. ^[1]

  Model weights are available on [Hugging Face](https://huggingface.co/inclusionAI/Ling-3.0-flash-Sante) and [ModelScope](https://www.modelscope.cn/models/inclusionAI/Ling-3.0-flash-Sante).

• The quickstart documents serving the model with SGLang or vLLM through an OpenAI-compatible API. ^[1]
• The model is a research model, not a medical device, and must not substitute for professional medical advice, diagnosis, or treatment. ^[1]

  Its outputs may be incomplete or incorrect and must not be used as a substitute for professional medical advice, diagnosis, or treatment.

• A speech model adds multilingual transcription and translation Daily brief 2026-10-26