SM

SmallThinker 4BA0.6B Instruct

SmallthinkerGGUF by TiinyUpdated 2025-08-05
Download on Hugging Face
Parameters4.27B
Context32K
Layers32
Architecturesmallthinker
Smallest quant1.9 GB
Quick answer

SmallThinker 4BA0.6B Instruct needs about 3.3 GB of memory at IQ4_NL (a 2.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 13 quants, from 1.9 GB (Q3_K_S) to 8.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.1 GB.

SmallThinker 4BA0.6B Instruct – a lightweight, on‑device Mixture‑of‑Experts LLM

SmallThinker 4BA0.6B Instruct is a 4.27B‑parameter model designed for local deployment. It uses a two‑level sparse MoE architecture and NoPE‑RoPE attention to reduce memory and compute. The model is available in multiple GGUF quantizations for llama.cpp.

Good at

  • Efficient on‑device inference
  • Low memory footprint
  • Supports 32768‑token context

Limits

  • Requires 8k context for memory estimates
  • No vision support
  • Apache‑2.0 license limits commercial use without attribution

Best for

  • Local chat assistants
  • Code generation on CPUs
  • Educational tools

Who it’s for. Developers and hobbyists who need a private LLM on modest hardware. The model runs well on consumer CPUs with 8k context and can be offloaded to storage with PowerInfer.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q3_K_S1.9 GB 2.7 GB
Fits Get file →
Q3_K2.1 GB 2.9 GB
Fits Get file →
Q4_02.3 GB 3.1 GB
Fits Get file →
IQ4_XS2.4 GB 3.1 GB
Fits Get file →
Q4_02.5 GB 3.2 GB
Fits Get file →
Q4_K_S2.5 GB 3.3 GB
Fits Get file →
IQ4_NL2.5 GB 3.3 GB
Fits Get file →
Q4_K2.6 GB 3.4 GB
Fits Get file →
Q4_12.7 GB 3.5 GB
Fits Get file →
Q5_K3.1 GB 3.8 GB
Fits Get file →
Q6_K3.5 GB 4.3 GB
Fits Get file →
Q8_04.5 GB 5.3 GB
Fits Get file →
F168.5 GB 9.3 GB
Too big Get file →

For 8GB VRAM, use Q4_0 (file size 2.29 GB, memory 3.1 GB). For 4GB VRAM, use Q3_K_S (file size 1.94 GB, memory 2.7 GB). For the smallest memory footprint, use Q3_K_S or Q3_K. All estimates are approximate.

What GPU do you need for SmallThinker 4BA0.6B 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 Q5_K (3.1 GB file, about 3.8 GB in use).
  • 6 GB VRAM: the best quant that fits is Q8_0 (4.5 GB file, about 5.3 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (4.5 GB file, about 5.3 GB in use).
  • 12 GB VRAM: the best quant that fits is F16 (8.5 GB file, about 9.3 GB in use).

Which SmallThinker 4BA0.6B Instruct quant should you choose?

IQ4_NL (2.5 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.

If memory is tight, the smallest file is Q3_K_S at 1.9 GB, with the quality trade-offs that come with fewer bits. A reasonable choice when a 4-bit file will not fit, mostly on larger models. Quality loss is noticeable but often acceptable.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturesmallthinker
Parameters4.27B (approximate, mixture of experts)
Layers32
Trained context32,768 tokens
KV cache (fp16)32 KB per token

Source

Quantized byTiiny
Licenseapache-2.0
Downloads (HF)1,326
Last updated2025-08-05

Frequently asked questions about SmallThinker 4BA0.6B Instruct

How much VRAM does SmallThinker 4BA0.6B Instruct need?

About 3.3 GB at IQ4_NL with an 8K context window. That is the 2.5 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 SmallThinker 4BA0.6B Instruct on an 8 GB GPU?

Yes, with Q8_0: the 4.5 GB file needs about 5.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of SmallThinker 4BA0.6B Instruct GGUF?

It depends on the quant: Q3_K_S is 1.9 GB, Q3_K is 2.1 GB, Q4_0 is 2.3 GB, IQ4_XS is 2.4 GB, Q4_0 is 2.5 GB, Q4_K_S is 2.5 GB and 7 more.

What context length does SmallThinker 4BA0.6B Instruct support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 0.3 GB at 8K and 1.1 GB at 32K.

Where can I download SmallThinker 4BA0.6B Instruct GGUF?

The files are on Hugging Face in the Tiiny/SmallThinker-4BA0.6B-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 SmallThinker 4BA0.6B 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 SmallThinker 4BA0.6B Instruct on 8GB VRAM?

Yes, the Q4_0 quantization requires about 3.1 GB of VRAM at 8k context, which fits within 8GB.

Does the model support vision tasks?

No, SmallThinker is a text‑only model and does not include vision capabilities.

What license applies to this model?

The model is released under the Apache‑2.0 license, which allows free use with attribution.

Other GGUFs of SmallThinker-4BA0.6B-Instruct

Similar models

Browse all →

More from Tiiny

All by Tiiny →

More Smallthinker models

All Smallthinker →

Other small models

All small models →