Model reference · open weights
Intern-S1-Pro is an open-weight language model from internlm, listed in the AxForge catalogue. AxForge can bring it up on EU-owned hardware for you on request — with the licence handled where one is required.
About
Intern-S1-Pro 💻Github Repo • 🤗Model Collections • 📜Technical Report • 💬Online Chat 👋 join us on <a href="https://discord.gg/xa29JuW87d" target="blank"Discord</a and <a href="https://cdn.vansin.top/intern-s1.jpg" target="blank"WeChat</a Introduction We introduce Intern-S1-Pro, a trillion-scale MoE multimodal scientific reasoning model. Intern-S1-Pro scales to 1T total parameters with 512 experts, activating 8 experts per token (22B activated parameters). The model delivers top-tier performance on advanced reasoning benchmarks and achieves leading results across key AI4Science domains (chemistry, materials, life-science, earth, etc.), while maintaining strong general multimodal and text capabilities. Features - State-of-the-art scientific reasoning, competitive with leading closed-source models across AI4Science tasks. - Strong general multimodal performance on various benchmarks. - Trillion-scale MoE training efficiency with STE routing (dense gradient for router training) and grouped routing for stable convergence and balanced expert parallelism. - Fourier Position Encoding (FoPE) + upgraded time-series modeling for better physical signal representation; supports long, heterogeneous time-series (10^0–10^6 points). Performance We evaluate the Intern-S1-Pro on various benchmarks, including general datasets and scientific datasets. We report the performance comparison with the recent VLMs and LLMs below. Note: <uUnderline</u means the best performance among open-sourced models, Bold indicates the best performance among all models. We use the OpenCompass and VLMEvalKit to evaluate all models. Quick Start Sampling Parameters We recommend using the following hyperparameters to ensure better results Serving [!IMPORTANT] Running a trillion-parameter model using the native Hugging Face forward method is challenging. We strongly recommend using an LLM inference engine (such as LMDeploy, vLLM, or SGLang) to host Intern-S1-Pro and accessing the model via API. Intern-S1-Pro can be deployed using any of the following LLM inference frameworks: - LMDeploy - vLLM - SGLang Detailed deployment examples for these frameworks are available in the Model Deployment Guide. Advanced Usag
Summarised from the published model card. Read the full card on the HuggingFace links below.
Specifications
| Maker | internlm |
|---|---|
| Type | Language models |
| Variants | 2 |
| Runs with | transformers |
| Released | 2026-02-02 |
| Popularity | 76k downloads / month |
| Likes | 278 |
| Licence | Open weights |
How it works
Variants
Open weights ship in several sizes and precisions. One page, all the variants — pick the one that fits your GPU. VRAM figures are estimates from model size.
Using it via the API
Once AxForge deploys intern-s1-pro for you, it answers on the OpenAI-compatible API — the same base URL and keys as every other model. (intern-s1-pro below is illustrative; you get the exact model name on deployment.)
$ curl -sS https://api.axforge.ai/v1/chat/completions \
-H "Authorization: Bearer $AXFORGE_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"intern-s1-pro","messages":[{"role":"user","content":"Hello"}]}'
Licence
Open weights under apache-2.0 — commercial use is permitted. Deploy it on AxForge EU hardware on request. Read the licence ↗