SM

SmolLM3 3B heretic

SmolLMGGUF by saidutta69Updated 2026-10-28
Download on Hugging Face
Parameters3.08B
Context64K
Layers36
Architecturesmollm3
Smallest quant1.3 GB
Quick answer

SmolLM3 3B heretic needs about 3.0 GB of memory at Q4_K_M (a 1.9 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 15 quants, from 1.3 GB (Q2_K) to 6.2 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 2.4 GB.

SmolLM3 3B heretic – a small, instruction‑tuned model for edge and CPU inference

SmolLM3 3B heretic is a 3.08 B parameter model with a 65536‑token context window. It is a decensored variant produced with Heretic abliteration, suppressing refusal behavior while preserving base model knowledge. The model runs natively in llama.cpp, Ollama, LM Studio, Jan, vLLM, and SGLang.

Good at

  • Reasoning
  • Roleplay
  • Tool usage

Limits

  • No vision support
  • Requires at least 2.4 GB VRAM for Q2_K
  • License is Apache‑2.0

Best for

  • Edge device inference
  • CPU‑only deployment
  • Low‑VRAM gaming PCs

Who it’s for. Developers needing a tiny instruction‑tuned model that answers directly instead of refusing. Ideal for on‑device inference or any use case blocked by RLHF‑era over‑refusal.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K1.3 GB 2.4 GB
Fits Get file →
Q3_K_S1.4 GB 2.5 GB
Fits Get file →
IQ3_S1.4 GB 2.5 GB
Fits Get file →
Q3_K_M1.6 GB 2.7 GB
Fits Get file →
Q3_K_L1.7 GB 2.8 GB
Fits Get file →
IQ4_XS1.7 GB 2.8 GB
Fits Get file →
Q4_01.8 GB 2.9 GB
Fits Get file →
Q4_K_S1.8 GB 2.9 GB
Fits Get file →
Q4_K_M1.9 GB 3.0 GB
Fits Get file →
Q4_12.0 GB 3.1 GB
Fits Get file →
Q5_K_S2.2 GB 3.3 GB
Fits Get file →
Q5_K_M2.2 GB 3.3 GB
Fits Get file →
Q6_K2.5 GB 3.6 GB
Fits Get file →
Q8_03.3 GB 4.4 GB
Fits Get file →
F166.2 GB 7.3 GB
Tight Get file →

For 8 GB VRAM, Q4_K_M (1.92 GB file, ~3 GB memory) is a good balance. If VRAM is 6 GB, use Q4_K_S (1.82 GB file, ~2.9 GB memory). For 4 GB VRAM, Q3_K_S (1.43 GB file, ~2.5 GB memory) is recommended. Memory estimates are approximate.

What GPU do you need for SmolLM3 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 Q6_K (2.5 GB file, about 3.6 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (3.3 GB file, about 4.4 GB in use).
  • 8 GB VRAM: the best quant that fits is F16 (6.2 GB file, about 7.3 GB in use).

Which SmolLM3 3B heretic quant should you choose?

Q4_K_M is the usual starting point for SmolLM3 3B heretic: it balances file size and quality well. Here it is a 1.9 GB file.

The Q8_0 file is 3.3 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 Q2_K at 1.3 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturesmollm3
Parameters3.08B
Layers36
Trained context65,536 tokens
KV cache (fp16)72 KB per token

Source

Quantized bysaidutta69
FamilySmolLM
Licenseapache-2.0
Downloads (HF)1,242
Last updated2026-10-28

Frequently asked questions about SmolLM3 3B heretic

How much VRAM does SmolLM3 3B heretic need?

About 3.0 GB at Q4_K_M with an 8K context window. That is the 1.9 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 SmolLM3 3B heretic on an 8 GB GPU?

Yes, with F16: the 6.2 GB file needs about 7.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of SmolLM3 3B heretic GGUF?

It depends on the quant: Q2_K is 1.3 GB, Q3_K_S is 1.4 GB, IQ3_S is 1.4 GB, Q3_K_M is 1.6 GB, Q3_K_L is 1.7 GB, IQ4_XS is 1.7 GB and 9 more.

What context length does SmolLM3 3B heretic support?

The GGUF header lists a trained context length of 65,536 tokens (64K). The KV cache grows with the context you use: about 0.6 GB at 8K and 2.4 GB at 32K.

Where can I download SmolLM3 3B heretic GGUF?

The files are on Hugging Face in the saidutta69/SmolLM3-3B-heretic repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does SmolLM3 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 SmolLM3 3B heretic on 8 GB VRAM?

Yes, Q4_K_M or Q4_K_S quants fit comfortably on 8 GB VRAM, using about 3 GB of memory at 8k context.

Does this model support vision tasks?

No, the model has no vision capabilities.

Is the model licensed for commercial use?

It is released under the Apache‑2.0 license, which allows commercial use.

Similar models

Browse all →

More from saidutta69

All by saidutta69 →

More SmolLM models

All SmolLM →

Other tiny models

All tiny models →