GR

Granite 4.0 H Small

GraniteGGUF by bartowskiUpdated 2025-10-02
Download on Hugging Face
Parameters32.2B
Context1024K
Layers40
Architecturegranitehybrid
Smallest quant0.1 GB
Quick answer

Granite 4.0 H Small needs about 25.6 GB of memory at Q4_K_M (a 19.7 GB file) with an 8K context window, so it fits on a 32 GB GPU. This repository has 28 quants, from 0.1 GB to 64.4 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 21.5 GB.

Granite 4.0 H Small – 32B parameter long‑context instruction model

Granite 4.0 H Small is a 32.2B parameter model with a 1,048,576 token context. It is fine‑tuned for instruction following, tool‑calling, and multilingual dialogue. The model is released under Apache 2.0 and supports English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese.

Good at

  • Long‑context instruction following
  • Tool‑calling and function integration
  • Multilingual dialogue support

Limits

  • Requires large VRAM for higher‑quality quants
  • No vision or audio capabilities
  • License is Apache 2.0

Best for

  • Enterprise AI assistants
  • Retrieval‑augmented generation
  • Code completion in multiple languages

Who it’s for. Developers building AI assistants that need robust instruction following and tool‑calling. The model is suitable for businesses that require multilingual support and long‑context reasoning.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.1 GB 6.0 GB
Fits Get file →
IQ2_XXS8.1 GB 14.0 GB
Too big Get file →
IQ2_XS9.3 GB 15.2 GB
Too big Get file →
IQ2_S9.3 GB 15.2 GB
Too big Get file →
IQ2_M10.5 GB 16.3 GB
Too big Get file →
Q2_K11.7 GB 17.6 GB
Too big Get file →
Q2_K_L11.8 GB 17.7 GB
Too big Get file →
IQ3_XXS13.1 GB 19.0 GB
Too big Get file →
IQ3_XS13.8 GB 19.6 GB
Too big Get file →
Q3_K_S14.4 GB 20.3 GB
Too big Get file →
IQ3_M15.0 GB 20.9 GB
Too big Get file →
Q3_K_M15.0 GB 20.9 GB
Too big Get file →
Q3_K_L15.6 GB 21.4 GB
Too big Get file →
Q3_K_XL15.7 GB 21.5 GB
Too big Get file →
IQ4_XS17.6 GB 23.4 GB
Too big Get file →
IQ4_NL18.5 GB 24.4 GB
Too big Get file →
Q4_018.8 GB 24.7 GB
Too big Get file →
Q4_K_S19.1 GB 25.0 GB
Too big Get file →
Q4_K_M19.7 GB 25.6 GB
Too big Get file →
Q4_K_L19.8 GB 25.7 GB
Too big Get file →
Q4_120.5 GB 26.3 GB
Too big Get file →
Q5_K_S22.4 GB 28.3 GB
Too big Get file →
Q5_K_M23.1 GB 29.0 GB
Too big Get file →
Q5_K_L23.2 GB 29.1 GB
Too big Get file →
Q6_K26.7 GB 32.5 GB
Too big Get file →
Q6_K_L26.8 GB 32.6 GB
Too big Get file →
Q8_034.3 GB 40.1 GB
Too big Get file →
BF1664.4 GB (2 parts) 70.3 GB
Too big Get file →

For 8 GB VRAM, use IQ2_XXS (6 GB memory at 8k context). With 16 GB VRAM, Q4_K_M (25.6 GB memory at 8k context) is a good balance of quality and size. For 32 GB VRAM, Q6_K (32.5 GB memory at 8k context) offers higher precision.

What GPU do you need for Granite 4.0 H Small?

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: no quant fits at an 8K context.
  • 6 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 6.0 GB in use).
  • 8 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 6.0 GB in use).
  • 12 GB VRAM: the best quant that fits is the main file (0.1 GB file, about 6.0 GB in use).
  • 16 GB VRAM: the best quant that fits is IQ2_S (9.3 GB file, about 15.2 GB in use).
  • 24 GB VRAM: the best quant that fits is IQ4_XS (17.6 GB file, about 23.4 GB in use).
  • 32 GB VRAM: the best quant that fits is Q5_K_L (23.2 GB file, about 29.1 GB in use).
  • 48 GB VRAM: the best quant that fits is Q8_0 (34.3 GB file, about 40.1 GB in use).

Which Granite 4.0 H Small quant should you choose?

Q4_K_M is the usual starting point for Granite 4.0 H Small: it balances file size and quality well. Here it is a 19.7 GB file.

The Q8_0 file is 34.3 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 0.1 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

Architecturegranitehybrid
Parameters32.2B (approximate, mixture of experts)
Layers40
Trained context1,048,576 tokens
KV cache (fp16)640 KB per token

Source

Quantized bybartowski
FamilyGranite
Licenseapache-2.0
Downloads (HF)2,855
Last updated2025-10-02

Frequently asked questions about Granite 4.0 H Small

How much VRAM does Granite 4.0 H Small need?

About 25.6 GB at Q4_K_M with an 8K context window. That is the 19.7 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 Granite 4.0 H Small on an 8 GB GPU?

Yes, with the smaller files: the 0.1 GB file needs about 6.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Granite 4.0 H Small GGUF?

It depends on the quant: file is 0.1 GB, IQ2_XXS is 8.1 GB, IQ2_XS is 9.3 GB, IQ2_S is 9.3 GB, IQ2_M is 10.5 GB, Q2_K is 11.7 GB and 22 more.

What context length does Granite 4.0 H Small support?

The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 5.4 GB at 8K and 21.5 GB at 32K.

Where can I download Granite 4.0 H Small GGUF?

The files are on Hugging Face in the bartowski/ibm-granite_granite-4.0-h-small-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 Granite 4.0 H Small 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 Granite 4.0 H Small on 8 GB VRAM?

Use the IQ2_XXS quant, which requires approximately 6 GB of VRAM at 8k context.

Does the model support vision or audio input?

No, the model only processes text and does not have vision or audio capabilities.

What license applies to this model?

The model is released under the Apache 2.0 license.

Other GGUFs of granite-4.0-h-small

Similar models

Browse all →

More from bartowski

All by bartowski →

More Granite models

All Granite →

Other medium models

All medium models →