Intern-S1
241B-total (28B-active) scientific multimodal MoE continually pretrained on 5T tokens (over 2.5T scientific) with Mixture-of-Rewards RL on 1,000+ tasks.
A specialist generalist for chemistry, materials and other science modalities (molecules, time series), post-trained in InternBootCamp. Apache-2.0 weights; paper arXiv 2508.15763 (2025-08-21).
- Date
- Thursday, 24 July 2025
- Lab
- Shanghai AI Laboratory (InternLM)
- Kind
- open-weights
- Access
- open weights
Figures
| Measure | Value | Measured by |
|---|---|---|
| Total / active parameters | 241B / 28B | company |
| Continued pretraining tokens | 5T over 2.5T scientific | company |
Date from the HF weights upload (2025-07-24 UTC); the lab's WAIC launch was around then but I did not open a dated post. Benchmark claims are the authors'.
Sources
This record was checked against its sources on 6 October 2026. How we check