Intern S2 397B — Hardware Requirements & GPU Compatibility
Vision
Intern-S2-397B is Shanghai AI Laboratory's multimodal foundation model, built for scientific intelligence and long-horizon agent tasks. It is a Mixture-of-Experts model with roughly 397 billion parameters, using a vision-language pretraining approach that learns directly from raw pages of scientific literature, jointly modeling text and visual relationships without an intermediate parsing step. It supports a thinking mode for deeper reasoning, enabled by default, and was trained with reinforcement learning across more than 20 scientific domains. It was evaluated with up to a 256K token context for text reasoning and 64K for multimodal inputs, and is released under the Apache 2.0 license. At roughly 397 billion parameters, running it locally needs multi-GPU server-class hardware; most users will use a hosted endpoint instead.
619 downloads
46 likesSpecifications
- Publisher
- InternLM
- Family
- Intern-S2
- Parameters
- 403.4B
- Architecture
- Qwen3_5MoeForConditionalGeneration
- Context Length
- 262,144 tokens
- Vocabulary Size
- 251,392
- Release Date
- 2026-09-13
- License
- Apache 2.0
Get Started
HuggingFace
How Much VRAM Does Intern S2 397B Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 171.9 GB | 187.9 GB | 171.45 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 176.9 GB | 192.9 GB | 176.50 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 197.1 GB | 213.1 GB | 196.67 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 202.1 GB | 218.1 GB | 201.71 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 242.5 GB | 258.5 GB | 242.05 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 287.9 GB | 303.9 GB | 287.44 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 333.3 GB | 349.2 GB | 332.82 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 403.9 GB | 419.8 GB | 403.42 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Intern S2 397B?
Q4_K_M · 242.5 GBIntern S2 397B (Q4_K_M) requires 242.5 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 316+ GB is recommended. Using the full 262K context window can add up to 16.0 GB, bringing total usage to 258.5 GB. No single GPU has enough memory — multi-GPU or cluster setups are needed.
Which Devices Can Run Intern S2 397B?
Q4_K_M · 242.5 GB3 devices with unified memory can run Intern S2 397B, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomWhere to Download Intern S2 397B
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Related Models
Frequently Asked Questions
- How much VRAM does Intern S2 397B need?
Intern S2 397B requires 242.5 GB of VRAM at Q4_K_M, or 807.3 GB at BF16. Full 262K context adds up to 16.0 GB (258.5 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 403.4B × 4.8 bits ÷ 8 = 242.1 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 16.4 GB (at full 262K context)
VRAM usage by quantization
Q4_K_M242.5 GBQ4_K_M + full context258.5 GB- Can NVIDIA GeForce RTX 5090 run Intern S2 397B?
No — Intern S2 397B requires at least 111.4 GB at IQ2_XXS, which exceeds the NVIDIA GeForce RTX 5090's 32 GB of VRAM.
- What's the best quantization for Intern S2 397B?
For Intern S2 397B, Q4_K_M (242.5 GB) offers the best balance of quality and VRAM usage. Q5_K_S (277.8 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 111.4 GB.
VRAM requirement by quantization
IQ2_XXS111.4 GBIQ3_XS166.8 GBQ3_K_L207.2 GBQ4_K_M ★242.5 GBQ5_K_S277.8 GBBF16807.3 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Intern S2 397B on a Mac?
Intern S2 397B requires at least 111.4 GB at IQ2_XXS, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Intern S2 397B locally?
Yes — Intern S2 397B can run locally on consumer hardware. At Q4_K_M quantization it needs 242.5 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Intern S2 397B?
At Q4_K_M, Intern S2 397B can reach ~66 tok/s on AMD Instinct MI350X. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B300 → 8000 ÷ 242.5 × 0.65 = ~163 tok/s
Estimated speed at Q4_K_M (242.5 GB)
~163 tok/s~66 tok/s~66 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Intern S2 397B?
At Q4_K_M, the download is about 242.05 GB. The full-precision BF16 version is 806.85 GB. The smallest option (IQ2_XXS) is 110.94 GB.
- Which GPUs can run Intern S2 397B?
No single consumer GPU has enough VRAM to run Intern S2 397B at Q4_K_M (242.5 GB). Multi-GPU or professional hardware is required.
- Which devices can run Intern S2 397B?
4 devices with unified memory can run Intern S2 397B at Q4_K_M (242.5 GB), including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), NVIDIA DGX A100 640GB, NVIDIA DGX H100. Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.