IN

Intern S2 397B

InternLMGGUF by bartowskiUpdated 2026-09-14
Download on Hugging Face
Parameters403B
Context256K
Layers61
Architectureqwen35moe
Smallest quant1.1 GB
Quick answer

Intern S2 397B needs about 256.1 GB of memory at Q4_K_M (a 254.6 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 25 quants, from 1.1 GB to 428.6 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.1 GB.

Intern S2 397B – a large multimodal foundation model

Intern S2 397B is a 403B‑parameter multimodal model from the InternLM family. It supports text and image input and is trained with a 262144‑token context window. The model is released under the Apache‑2.0 license.

Good at

  • Long‑context reasoning
  • Multimodal vision‑language tasks
  • Scientific problem solving

Limits

  • Requires large VRAM for full precision
  • No native support for audio or video
  • Limited to 262144‑token context

Best for

  • Scientific literature analysis
  • Long‑form content generation
  • Agentic task planning

Who it’s for. Developers building scientific or long‑horizon AI applications that need both text and image understanding. The model is suitable for research and production use where a large context window is required.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—1.1 GB 2.7 GB
Fits Get file →
IQ1_S91.6 GB (3 parts) 93.1 GB
Too big Get file →
IQ1_M101.8 GB (3 parts) 103.3 GB
Too big Get file →
IQ2_XXS118.8 GB (3 parts) 120.4 GB
Too big Get file →
IQ2_XS125.2 GB (4 parts) 126.7 GB
Too big Get file →
IQ2_S131.6 GB (4 parts) 133.1 GB
Too big Get file →
IQ2_M146.1 GB (4 parts) 147.6 GB
Too big Get file →
Q2_K154.5 GB (4 parts) 156.0 GB
Too big Get file →
IQ3_XXS173.3 GB (5 parts) 174.8 GB
Too big Get file →
IQ3_XS181.8 GB (5 parts) 183.4 GB
Too big Get file →
Q3_K_S183.8 GB (5 parts) 185.3 GB
Too big Get file →
Q3_K_M193.7 GB (5 parts) 195.2 GB
Too big Get file →
Q3_K_L203.9 GB (6 parts) 205.5 GB
Too big Get file →
IQ3_M213.2 GB (6 parts) 214.7 GB
Too big Get file →
IQ4_XS223.3 GB (6 parts) 224.8 GB
Too big Get file →
Q4_0229.2 GB (6 parts) 230.7 GB
Too big Get file →
Q4_K_S237.2 GB (7 parts) 238.8 GB
Too big Get file →
Q4_1252.8 GB (7 parts) 254.4 GB
Too big Get file →
IQ4_NL253.3 GB (7 parts) 254.8 GB
Too big Get file →
Q4_K_M254.6 GB (7 parts) 256.1 GB
Too big Get file →
Q5_K_S286.6 GB (8 parts) 288.1 GB
Too big Get file →
Q5_K_M309.4 GB (8 parts) 310.9 GB
Too big Get file →
Q6_K_S336.9 GB (9 parts) 338.4 GB
Too big Get file →
Q6_K353.5 GB (9 parts) 355.1 GB
Too big Get file →
Q8_0428.6 GB (11 parts) 430.1 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (254.58 GB file, ~256.1 GB memory at 8 k context). For 16 GB VRAM, Q5_K_M (309.36 GB file, ~310.9 GB memory) offers a balance of size and performance. For 32 GB VRAM, Q6_K (353.55 GB file, ~355.1 GB memory) provides higher fidelity.

What GPU do you need for Intern S2 397B?

These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.

  • 4 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 6 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 16 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 24 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 32 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).
  • 48 GB VRAM: the best quant that fits is the main file (1.1 GB file, about 2.7 GB in use).

Which Intern S2 397B quant should you choose?

Q4_K_M is the usual starting point for Intern S2 397B: it balances file size and quality well. Here it is a 254.6 GB file.

The Q8_0 file is 428.6 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.

If memory is tight, the smallest file is the first one in the table at 1.1 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureqwen35moe
Parameters403B (approximate, mixture of experts)
Layers61
Trained context262,144 tokens
KV cache (fp16)122 KB per token

Source

Quantized bybartowski
FamilyInternLM
Licenseapache-2.0
Downloads (HF)2,450
Last updated2026-09-14

Frequently asked questions about Intern S2 397B

How much VRAM does Intern S2 397B need?

About 256.1 GB at Q4_K_M with an 8K context window. That is the 254.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.

Can I run Intern S2 397B on an 8 GB GPU?

Yes, with the smaller files: the 1.1 GB file needs about 2.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Intern S2 397B GGUF?

It depends on the quant: file is 1.1 GB, IQ1_S is 91.6 GB, IQ1_M is 101.8 GB, IQ2_XXS is 118.8 GB, IQ2_XS is 125.2 GB, IQ2_S is 131.6 GB and 19 more.

What context length does Intern S2 397B support?

The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.0 GB at 8K and 4.1 GB at 32K.

Where can I download Intern S2 397B GGUF?

The files are on Hugging Face in the bartowski/Intern-S2-397B-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does Intern S2 397B use?

Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Can I run Intern S2 397B on 8 GB VRAM?

Use the Q4_K_M quant, which requires about 256.1 GB of VRAM at 8 k context. It is the smallest quant that supports the full model.

Does the model support image input?

Yes. The model includes multimodal projector files that pair with any quant for image processing.

What license applies to this model?

The model is released under the Apache‑2.0 license.

Other GGUFs of Intern-S2-397B

Similar models

Browse all →

More from bartowski

All by bartowski →

More InternLM models

All InternLM →

Other large models

All large models →