Merlin
Stanford's 3D vision-language model for abdominal CT, pretrained on scans, EHR codes and radiology reports for classification, retrieval and report generation.
| Category | Clinical & biomedical LLMs |
|---|---|
| GitHub | StanfordMIMI/Merlin |
| First public | 2024-06-10 (source) |
| Paper | https://www.nature.com/articles/s41586-026-10181-8 · Nature (2026) |
| License | MIT |
| Pricing | Open source |
| Tags | #ct#vision-language#radiology |
| Last checked | 2026-10-16 |
Specs
| Modality | 3D CT + EHR + reports |
|---|---|
| Base model | I3D ResNet image encoder + Clinical-Longformer text encoder |
| Parameters | — |
| GitHub stars | 490 |
| Interface | — |
Specs come from each project's README, paper or official page · Merlin README · Merlin paper · Merlin source 3. “—” means we couldn't verify it.
Compare with…
Pick 1–3 other Clinical & biomedical LLMs tools to compare side by side with Merlin.
Related
Segment Anything adapted to medical images: promptable segmentation across CT, MRI, ultrasound, endoscopy and more.
Clinical & biomedical LLMsOpen source#segmentation#foundation-model
Biomedical vision-language assistant trained on PubMed Central figure-caption data.
Clinical & biomedical LLMsOpen source (non-commercial)#vlm#medical-imaging
Open medical LLMs (7B/70B) adapted from Llama 2 on curated medical corpora.
Clinical & biomedical LLMsOpen source#llm#medical