Intern-S1
Intern-S1 is a large-scale multimodal foundation model developed by Shanghai AI Laboratory. It boasts an extremely large scale of approximately 2.4 trillion parameters and is equipped with a 128K context window.
Parameters
2410.0B
Context Window
128K
License
Apache 2.0
Release Date
2025-07-27
API Pricing
API pricing for this model is not yet available
Strengths
- ・Giant parameter scale over 2.4 trillion
- ・Achieves advanced multimodal processing
- ・Wide 128K context window
Weaknesses
- ・Massive model size of 481.6GB
- ・Requires extremely high computational resources
- ・Increased inference cost due to enormous scale
Use Cases
- ・Advanced multimodal data analysis
- ・Long-document understanding at scale
- ・Complex multimodal tasks
Deep Analysis
Architecture
Multimodal Foundation Model
Scientific multimodal model
Specialization
Science (Chem, Bio, Math, Physics)
Multi-domain scientific reasoning
Release Date
July 2025
License
Apache 2.0
GitHub Stars
772
InternLM/Intern-S1
Key Claim
#1 open-source multimodal
Multimodal capability ranking
Strengths
- ・#1 open-source multimodal capability globally
- ・Strong scientific reasoning across chemistry, biology, math, physics
- ・Apache 2.0 license
- ・Competitive general capabilities alongside specialized science
- ・From Shanghai AI Lab (strong research pedigree)
Weaknesses
- ・Specialized for scientific tasks
- ・Limited general-purpose benchmarks published
- ・Chinese model ecosystem
Competitor Comparison
| Model | Arena | SWE | GPQA | Price |
|---|---|---|---|---|
| InternLM3 | - | - | - | Free (open) |
| Qwen 2.5 | - | - | - | Free (open) |
| GPT-4o | - | - | - | Paid |
Intern-S1 is Shanghai AI Lab's scientific multimodal foundation model, claiming #1 open-source multimodal capability globally. Released July 2025 under Apache 2.0, it excels in chemistry, biology, math, and physics while maintaining strong general capabilities.
Analysis generated: 2026-05-24