by inclusionAI
Open-weights 124B MoE healthcare model (5.1B active) from inclusionAI, specialized from Ling-3.0-flash.
Updated EU status checked Last shown
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Ling-3.0-flash-Santé is a healthcare-specialized language model from inclusionAI, built on Ling-3.0-flash. The name comes from the French word for "health". It has 124B total parameters with 5.1B active per token, using a hybrid-linear mixture-of-experts architecture that combines Kimi Delta Attention, gated MLA and sparse MoE layers. Training extends to a 256K-token context.
It targets medical reasoning, healthcare research, deep research, multi-agent collaboration and tool-assisted workflows, while retaining general capabilities. Thinking mode is on by default, and tool calling is supported through compatible serving frameworks. The model card reports evaluations on professional healthcare tasks, medical knowledge and safety, complex reasoning, BrowseComp, HLE with tools and tool-use benchmarks.
Weights are published on Hugging Face and ModelScope under the MIT license, and the model is also listed on Vercel AI Gateway. Self-hosting is documented with SGLang; the example uses four 141GB-class GPUs for a 256K context. The vendor states it is a research model, not a medical device, and its outputs must not replace professional medical advice. High-stakes use needs qualified human review.
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Open medical MoE model with 124B total and 5.1B active parameters, specialized from Ling 3 Flash. Competitive on clinical knowledge and medical reasoning benchmarks; full model is about 255 GB. The creator attributes it to Anthropic, which appears to be a misstatement.
Closest neighbours by meaning, not just by category.
Model · inclusionAI
Open-weights 560B hybrid-linear MoE reasoning model (25B active, 1M context) from inclusionAI, MIT licensed.
Model · Baidu
Baidu's open-weights 280B-parameter sparse MoE base model (14B active) with 1M-token context, released pre-post-training.
Model · Rednote
Open MoE language model from Rednote (Xiaohongshu) with 208B total and 16B active parameters.
Model · OpenAI
OpenAI's open-weight reasoning models, gpt-oss-120b and gpt-oss-20b, released under Apache 2.0.
Model · Alibaba
Alibaba's Qwen3.8 Flash Next: a 125B-parameter MoE language model with about 6B active parameters, previewing Qwen 4 design.
Model · Z.ai
Z.ai's open-weights 753B MoE language model, a post-trained GLM-5.2 aimed at coding, agentic tasks and cybersecurity.