462
▲ 47% vs 2024
publications on multimodal biomedical AI in 2025
20212025
What the chart shows
Multimodal foundation models combine types of data, such as medical images with text, or tissue images with genomic data. RAISE Health counted 462 publications on them for biomedical discovery in 2025, up 47% from 314 in 2024.
The field barely existed in 2021, with 2 publications, and jumped to 171 in 2023. The AI Index singles out vision-language models, which pair images with text, and vision-omics models, which link images to genomic or transcriptomic data.
Key numbers
| Measure | Value |
|---|---|
| 2025 | 462 |
| 2024 | 314 |
| 2023 | 171 |
| 2022 | 16 |
| 2021 | 2 |
How to read it
- PubMed publications identified by RAISE Health as being about multimodal foundation models for biomedical discovery.
- Counts measure research activity, not quality or clinical use.
Source: RAISE Health, 2026, via the AI Index Report 2026 (Stanford HAI), figure 6.1.13: Number of publications on multimodal biomedical AI, 2021–25. Chart and text by AI Adoption.
Numbers checked against the published data on 20 October 2026.
Download the data (CSV) · Cite this chart
Numbers checked against the published data on 20 October 2026.
Download the data (CSV) · Cite this chart
Discussion