AI BriefinginclusionAI releases Ling-3.0-flash-Sante, a 124B healthcare-specialized MoE model
·2026.10.23 00:18
Key point
The model features a hybrid-linear MoE architecture supporting 256K token contexts and is available on Hugging Face and ModelScope.
Details
- inclusionAI released Ling-3.0-flash-Sante, a healthcare-specialized model built on Ling-3.0-flash with 124B total parameters and 5.1B active parameters per token.
- The model utilizes a hybrid-linear MoE architecture combining Kimi Delta Attention (KDA), gated MLA, and sparse MoE layers to achieve efficient inference at scale.
- Training extends to 256K tokens, enabling long-document analysis, evidence synthesis, and multi-step research workflows.
- Weights are available on Hugging Face and ModelScope, with thinking mode enabled by default and tool calling supported through compatible serving frameworks.
- The release is designated as a research model, not a medical device, requiring qualified human review and compliance with healthcare regulations for high-stakes use.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
inclusionAI releases Ling-3.1-flash: A 560B-parameter hybrid reasoning model with 1M-token context
DeepSeek V4.1 Flash 1017 scores 45 on Artificial Analysis Intelligence Index