Ant Ling Open-Sources Ling-3.0-flash-Sante, a 124B Healthcare MoE Model
A source-linked abstract prepared once when this article was collected.
Ant Ling releases Ling-3.0-flash-Sante, a healthcare-specialized model built on Ling-3.0-flash with 124B total and 5.1B active parameters per token. It is designed for medical reasoning, healthcare research, deep research, and tool-assisted workflows. The architecture combines Kimi Delta Attention, gated MLA, and sparse MoE layers, with training extended to 256K tokens. Weights are available on Hugging Face and ModelScope. The model card states it is a research model, not a medical device, and outputs must not substitute for professional medical advice.
Key takeaways
- The model has 124B total parameters with 5.1B active per token, built on Ling-3.0-flash.
- Its hybrid architecture combines Kimi Delta Attention, gated MLA, and sparse MoE layers.
- Training extends to 256K tokens, supporting long-document analysis and multi-step research.
- The model card states it is a research model, not a medical device, and needs human review.
Why it matters
An open-weight, healthcare-specialized model with long context and tool use could enable research and retrieval workflows in medicine. The authors stress it must not be used for diagnosis or treatment without qualified oversight.
Read with care
Benchmark results are reported by the developer and the model card notes they do not guarantee real-world clinical performance; source claims not independently checked.
Discussion
Comments are reviewed before publication. Discuss the evidence, not the person.
No published comments yet.
Reader discussion opens with accounts. You can read the original source and published comments now.