GR

Granite 4.2 3b Heretic

GraniteGGUF by mradermacherUpdated 2026-09-05
Download on Hugging Face
Parameters3.66B
Context128K
Layers40
Architecturegranite
Smallest quant0.0 GB
Quick answer

Granite 4.2 3b Heretic needs about 3.4 GB of memory at Q4_K_M (a 2.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 25 quants, from 0.0 GB to 3.0 GB (Q6_K). Longer context needs more: at 32K tokens the KV cache alone adds about 2.7 GB.

Granite 4.2 3b Heretic – a 3.66B parameter causal LLM

Granite 4.2 3b Heretic is a 3.66B parameter causal language model based on the Granite architecture. It is a representation‑edited version of ibm‑granite/granite‑4.2‑3b, released under Apache 2.0. The model supports reasoning and tool usage.

Good at

  • Reasoning tasks
  • Tool invocation
  • Instruction following

Limits

  • No vision support
  • Large context requires significant VRAM
  • Not fully evaluated for safety

Best for

  • Research into representation editing
  • Evaluating refusal behavior
  • Local prototyping

Who it’s for. Researchers and developers who need a lightweight, editable Granite model for experimentation. The model is suitable for local inference on systems with moderate VRAM and for studying alignment trade‑offs.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 1.2 GB
Fits Get file →
IQ1_S0.9 GB 2.1 GB
Fits Get file →
IQ1_M1.0 GB 2.2 GB
Fits Get file →
IQ2_XXS1.1 GB 2.3 GB
Fits Get file →
IQ2_XS1.2 GB 2.4 GB
Fits Get file →
IQ2_S1.3 GB 2.4 GB
Fits Get file →
IQ2_M1.4 GB 2.5 GB
Fits Get file →
Q2_K_S1.4 GB 2.5 GB
Fits Get file →
Q2_K1.5 GB 2.6 GB
Fits Get file →
IQ3_XXS1.5 GB 2.7 GB
Fits Get file →
IQ3_XS1.6 GB 2.8 GB
Fits Get file →
Q3_K_S1.7 GB 2.8 GB
Fits Get file →
IQ3_S1.7 GB 2.9 GB
Fits Get file →
IQ3_M1.7 GB 2.9 GB
Fits Get file →
Q3_K_M1.8 GB 3.0 GB
Fits Get file →
Q3_K_L2.0 GB 3.1 GB
Fits Get file →
IQ4_XS2.0 GB 3.2 GB
Fits Get file →
IQ4_NL2.1 GB 3.3 GB
Fits Get file →
Q4_02.1 GB 3.3 GB
Fits Get file →
Q4_K_S2.1 GB 3.3 GB
Fits Get file →
Q4_K_M2.2 GB 3.4 GB
Fits Get file →
Q4_12.3 GB 3.5 GB
Fits Get file →
Q5_K_S2.6 GB 3.7 GB
Fits Get file →
Q5_K_M2.6 GB 3.8 GB
Fits Get file →
Q6_K3.0 GB 4.2 GB
Fits Get file →

For 8 GB VRAM, use Q4_K_M (2.24 GB file, ~3.4 GB VRAM at 8 k context). For 12 GB VRAM, Q5_K_M (2.61 GB file, ~3.8 GB VRAM) or Q6_K (3.01 GB file, ~4.2 GB VRAM). IQ1_S (0.93 GB file) is the smallest option but uses ~2.1 GB VRAM.

What GPU do you need for Granite 4.2 3b Heretic?

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 Q5_K_M (2.6 GB file, about 3.8 GB in use).
  • 6 GB VRAM: the best quant that fits is Q6_K (3.0 GB file, about 4.2 GB in use).

Which Granite 4.2 3b Heretic quant should you choose?

Q4_K_M is the usual starting point for Granite 4.2 3b Heretic: it balances file size and quality well. Here it is a 2.2 GB file.

If memory is tight, the smallest file is the first one in the table at 0.0 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

Architecturegranite
Parameters3.66B
Layers40
Trained context131,072 tokens
KV cache (fp16)80 KB per token

Source

Quantized bymradermacher
FamilyGranite
Licenseapache-2.0
Downloads (HF)487
Last updated2026-09-05

Frequently asked questions about Granite 4.2 3b Heretic

How much VRAM does Granite 4.2 3b Heretic need?

About 3.4 GB at Q4_K_M with an 8K context window. That is the 2.2 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 4.2 3b Heretic on an 8 GB GPU?

Yes, with Q6_K: the 3.0 GB file needs about 4.2 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Granite 4.2 3b Heretic GGUF?

It depends on the quant: file is 0.0 GB, IQ1_S is 0.9 GB, IQ1_M is 1.0 GB, IQ2_XXS is 1.1 GB, IQ2_XS is 1.2 GB, IQ2_S is 1.3 GB and 19 more.

What context length does Granite 4.2 3b Heretic support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.7 GB at 8K and 2.7 GB at 32K.

Where can I download Granite 4.2 3b Heretic GGUF?

The files are on Hugging Face in the mradermacher/granite-4.2-3b-Heretic-i1-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 4.2 3b Heretic 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 this model on 8 GB VRAM?

Yes, using Q4_K_M or smaller quants. The estimated VRAM usage is about 3.4 GB at 8 k context.

Does the model support image input?

No, the model has no vision capabilities.

What license applies?

Apache 2.0, subject to the base model's terms and conditions.

Other GGUFs of granite-4.2-3b-Heretic

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Granite models

All Granite →

Other small models

All small models →