GR

granite-3.2-8b-instruct

GraniteGGUF by MungertUpdated 2025-09-24
Download on Hugging Face
Parameters8.17B
Context128K
Layers40
Architecturegranite
Smallest quant1.8 GB
Quick answer

granite-3.2-8b-instruct needs about 6.7 GB of memory at Q4_K_M (a 4.9 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 35 quants, from 1.8 GB (TQ1_0) to 16.3 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 5.4 GB.

Granite-3.2-8B-Instruct, an 8.17B parameter long‑context instruction model

Granite-3.2-8B-Instruct is a 8.17B parameter model fine‑tuned for long‑context instruction following. It supports tool usage and is available under the Apache‑2.0 license. The model was generated with llama.cpp and offers multiple quantization options.

Good at

  • Long‑context instruction following
  • Tool‑calling tasks
  • Multilingual dialog

Limits

  • No vision support
  • Requires 8k context for memory estimates
  • Licensed under Apache‑2.0

Best for

  • Chat assistants needing long context
  • Code generation with tool integration
  • Multilingual support

Who it’s for. Developers building AI assistants or chatbots that require long‑context understanding and tool‑calling capabilities. The model is suitable for applications that can run on systems with sufficient VRAM and comply with the Apache‑2.0 license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
TQ1_01.8 GB 3.7 GB
Fits Get file →
TQ2_02.2 GB 4.1 GB
Fits Get file →
IQ2_XXS2.5 GB 4.3 GB
Fits Get file →
IQ2_XS2.7 GB 4.5 GB
Fits Get file →
IQ2_S2.8 GB 4.6 GB
Fits Get file →
IQ2_M2.9 GB 4.8 GB
Fits Get file →
Q2_K_S3.0 GB 4.9 GB
Fits Get file →
Q2_K_M3.1 GB 4.9 GB
Fits Get file →
IQ3_XXS3.3 GB 5.2 GB
Fits Get file →
IQ3_XS3.4 GB 5.3 GB
Fits Get file →
IQ3_M3.8 GB 5.7 GB
Fits Get file →
IQ3_S3.8 GB 5.7 GB
Fits Get file →
Q3_K_S3.9 GB 5.8 GB
Fits Get file →
Q3_K_M4.0 GB 5.8 GB
Fits Get file →
IQ4_XS4.4 GB 6.3 GB
Fits Get file →
Q4_04.6 GB 6.4 GB
Fits Get file →
IQ4_NL4.7 GB 6.5 GB
Fits Get file →
Q4_K_S4.7 GB 6.6 GB
Fits Get file →
Q4_K_M4.9 GB 6.7 GB
Fits Get file →
Q4_15.1 GB 7.0 GB
Fits Get file →
BF165.2 GB 7.0 GB
Fits Get file →
F165.2 GB 7.0 GB
Fits Get file →
Q5_05.6 GB 7.5 GB
Tight Get file →
Q5_K_S5.7 GB 7.6 GB
Tight Get file →
Q5_K_M5.8 GB 7.7 GB
Tight Get file →
Q5_16.1 GB 8.0 GB
Tight Get file →
Q6_K_M6.7 GB 8.5 GB
Too big Get file →
BF166.9 GB 8.8 GB
Too big Get file →
F166.9 GB 8.8 GB
Too big Get file →
Q8_08.7 GB 10.5 GB
Too big Get file →
BF168.9 GB 10.7 GB
Too big Get file →
F168.9 GB 10.7 GB
Too big Get file →
BF1611.6 GB 13.5 GB
Too big Get file →
F1611.6 GB 13.5 GB
Too big Get file →
BF1616.3 GB 18.2 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (file size 4.9 GB, estimated memory 6.7 GB at 8k context). For 12 GB VRAM, Q5_K_M (file size 5.82 GB, estimated memory 7.7 GB) is a good balance. If more VRAM is available, Q8_0 (file size 8.68 GB, estimated memory 10.5 GB) provides higher precision.

What GPU do you need for granite-3.2-8b-instruct?

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 TQ1_0 (1.8 GB file, about 3.7 GB in use).
  • 6 GB VRAM: the best quant that fits is Q3_K_M (4.0 GB file, about 5.8 GB in use).
  • 8 GB VRAM: the best quant that fits is Q5_1 (6.1 GB file, about 8.0 GB in use).
  • 12 GB VRAM: the best quant that fits is BF16 (8.9 GB file, about 10.7 GB in use).
  • 16 GB VRAM: the best quant that fits is BF16 (11.6 GB file, about 13.5 GB in use).
  • 24 GB VRAM: the best quant that fits is BF16 (16.3 GB file, about 18.2 GB in use).

Which granite-3.2-8b-instruct quant should you choose?

Q4_K_M is the usual starting point for granite-3.2-8b-instruct: it balances file size and quality well. Here it is a 4.9 GB file.

The Q8_0 file is 8.7 GB, about 1.8 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 TQ1_0 at 1.8 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturegranite
Parameters8.17B
Layers40
Trained context131,072 tokens
KV cache (fp16)160 KB per token

Source

Quantized byMungert
FamilyGranite
Licenseapache-2.0
Downloads (HF)713
Last updated2025-09-24

Frequently asked questions about granite-3.2-8b-instruct

How much VRAM does granite-3.2-8b-instruct need?

About 6.7 GB at Q4_K_M with an 8K context window. That is the 4.9 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-3.2-8b-instruct on an 8 GB GPU?

Yes, with Q5_1: the 6.1 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of granite-3.2-8b-instruct GGUF?

It depends on the quant: TQ1_0 is 1.8 GB, TQ2_0 is 2.2 GB, IQ2_XXS is 2.5 GB, IQ2_XS is 2.7 GB, IQ2_S is 2.8 GB, IQ2_M is 2.9 GB and 29 more.

What context length does granite-3.2-8b-instruct support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.3 GB at 8K and 5.4 GB at 32K.

Where can I download granite-3.2-8b-instruct GGUF?

The files are on Hugging Face in the Mungert/granite-3.2-8b-instruct-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-3.2-8b-instruct 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-3.2-8B-Instruct on 8 GB VRAM?

Yes, the Q4_K_M quantization requires approximately 6.7 GB of VRAM at 8k context, which fits within an 8 GB budget.

Does this model support vision tasks?

No, Granite-3.2-8B-Instruct does not include vision capabilities.

What license applies to this model?

The model is released under the Apache‑2.0 license.

Other GGUFs of granite-3.1-8b-instruct

Similar models

Browse all →

More from Mungert

All by Mungert →

More Granite models

All Granite →

Other small models

All small models →