SP

Abliterated

SparkGGUF by alexwkleungUpdated 2026-10-22
Download on Hugging Face
Parameters4.11B
Context1024K
Layers36
Architecturespark2_5
Smallest quant2.6 GB
Quick answer

Abliterated needs about 4.3 GB of memory at Q4_K_M (a 2.6 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 4 quants, from 2.6 GB (Q4_K_M) to 4.4 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.8 GB.

Abliterated Spark-X2.6-4B: 4.1B parameter model with reduced refusals

Abliterated is a 4.11B parameter Spark model. It is a dense, low/high effort thinking model. The build removes a refusal direction from the residual stream, reducing over‑refusal.

Good at

  • Reasoning
  • Tool use
  • Agentic workflows

Limits

  • No vision support
  • Requires 8k context for estimates
  • Apache‑2.0 license

Best for

  • On‑device agents
  • Software engineering tasks
  • Long‑context reasoning

Who it’s for. Developers building on‑device AI agents who need a compact model with reduced refusal behaviour. The model is suitable for applications that can run on GPUs with 8k context support.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M2.6 GB 4.3 GB
Fits Get file →
Q5_K_M3.0 GB 4.7 GB
Fits Get file →
Q6_K3.4 GB 5.1 GB
Fits Get file →
Q8_04.4 GB 6.1 GB
Fits Get file →

For 8 GB VRAM use Q4_K_M (file size 2.6 GB, estimated memory 4.3 GB). For 12 GB VRAM use Q5_K_M (2.98 GB, 4.7 GB). For 16 GB VRAM use Q6_K (3.38 GB, 5.1 GB).

What GPU do you need for Abliterated?

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 Q6_K (3.4 GB file, about 5.1 GB in use).
  • 8 GB VRAM: the best quant that fits is Q8_0 (4.4 GB file, about 6.1 GB in use).

Which Abliterated quant should you choose?

Q4_K_M is the usual starting point for Abliterated: it balances file size and quality well. Here it is a 2.6 GB file.

The Q8_0 file is 4.4 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.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturespark2_5
Parameters4.11B
Layers36
Trained context1,048,576 tokens
KV cache (fp16)144 KB per token

Source

Quantized byalexwkleung
FamilySpark
Licenseapache-2.0
Downloads (HF)1,874
Last updated2026-10-22

Frequently asked questions about Abliterated

How much VRAM does Abliterated need?

About 4.3 GB at Q4_K_M with an 8K context window. That is the 2.6 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 Abliterated on an 8 GB GPU?

Yes, with Q8_0: the 4.4 GB file needs about 6.1 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of Abliterated GGUF?

It depends on the quant: Q4_K_M is 2.6 GB, Q5_K_M is 3.0 GB, Q6_K is 3.4 GB, Q8_0 is 4.4 GB.

What context length does Abliterated 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 1.2 GB at 8K and 4.8 GB at 32K.

Where can I download Abliterated GGUF?

The files are on Hugging Face in the alexwkleung/Spark-X2.6-4B-Abliterated-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 Abliterated 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 Abliterated on 8 GB VRAM?

Yes, the Q4_K_M quantization uses about 4.3 GB at 8k context, fitting within 8 GB VRAM.

Does Abliterated support vision?

No, the model has no vision capabilities.

What is the license of Abliterated?

The model is released under the Apache‑2.0 license.

Other GGUFs of Spark-X2.6-4B

Similar models

Browse all →

More from alexwkleung

All by alexwkleung →

More Spark models

All Spark →

Other small models

All small models →