ZA

ZAC-1 decoder (keep2_plain, head step 50000)

ZacGGUF by ZyphraUpdated 2026-11-12
Download on Hugging Face
Parameters1.06B
Context—
Layers—
Architecturemossdual_istft
Smallest quant0.5 GB
Quick answer

ZAC-1 decoder (keep2_plain, head step 50000) needs about 2.2 GB of memory at F32 (a 1.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 4 quants, from 0.5 GB (F16) to 4.0 GB (F32).

ZAC-1 decoder (keep2_plain, head step 50000) is a 1.06B-parameter Zac model in GGUF format, quantized by Zyphra. The repository has 4 quant files, from 0.5 GB (F16) to 4.0 GB (F32).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
F160.5 GB 1.0 GB
Fits Get file →
F321.5 GB 2.2 GB
Fits Get file →
F162.2 GB 3.0 GB
Fits Get file →
F324.0 GB 5.0 GB
Fits Get file →

What GPU do you need for ZAC-1 decoder (keep2_plain, head step 50000)?

These figures add the file size, an fp16 KV cache at an 8K context window 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 F16 (2.2 GB file, about 3.0 GB in use).
  • 6 GB VRAM: the best quant that fits is F32 (4.0 GB file, about 5.0 GB in use).

Which ZAC-1 decoder (keep2_plain, head step 50000) quant should you choose?

F32 (1.5 GB) is the middle option in this repository. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.

If memory is tight, the smallest file is F16 at 0.5 GB, with the quality trade-offs that come with fewer bits. The unquantized model. It gives original quality and needs the most memory, so it is mainly for conversion or reference.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturemossdual_istft
Parameters1.06B

Source

Quantized byZyphra
FamilyZac
RepositoryZyphra/ZAC-1
Licenseapache-2.0
Downloads (HF)99
Last updated2026-11-12

Frequently asked questions about ZAC-1 decoder (keep2_plain, head step 50000)

How much VRAM does ZAC-1 decoder (keep2_plain, head step 50000) need?

About 2.2 GB at F32 with an 8K context window. That is the 1.5 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 ZAC-1 decoder (keep2_plain, head step 50000) on an 8 GB GPU?

Yes, with F32: the 4.0 GB file needs about 5.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of ZAC-1 decoder (keep2_plain, head step 50000) GGUF?

It depends on the quant: F16 is 0.5 GB, F32 is 1.5 GB, F16 is 2.2 GB, F32 is 4.0 GB.

Where can I download ZAC-1 decoder (keep2_plain, head step 50000) GGUF?

The files are on Hugging Face in the Zyphra/ZAC-1 repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

What license does ZAC-1 decoder (keep2_plain, head step 50000) 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.

Similar models

Browse all →

More from Zyphra

All by Zyphra →

Other tiny models

All tiny models →