LI

LightOnOCR 2 1B Ocr Soup

LightonocrGGUF by noctrexUpdated 2026-01-24
Download on Hugging Face
Parameters596M
Context16K
Layers28
Architectureqwen3
Smallest quant0.0 GB
Quick answer

LightOnOCR 2 1B Ocr Soup needs about 1.8 GB of memory at Q4_K_M (a 0.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 9 quants, from 0.0 GB to 1.2 GB (F16).

LightOnOCR‑2 1B OCR model for fast, accurate document text extraction

LightOnOCR‑2 1B Ocr Soup is a vision‑language model that converts PDFs, scans, and images into clean, naturally ordered text. It is 0.596 billion parameters and supports 16384‑token context. The model is licensed under Apache‑2.0.

Good at

  • Fast OCR on high‑resolution documents
  • Robust handling of LaTeX and scientific text
  • Efficient on modest hardware

Limits

  • No vision support beyond OCR
  • Requires 8 k context for estimates
  • No multilingual language list provided

Best for

  • Batch OCR of scanned PDFs
  • Extracting text from scientific articles
  • Processing receipts and forms

Who it’s for. Developers building OCR pipelines that need speed and accuracy on a single GPU. Researchers requiring a lightweight vision‑language model for document conversion.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 1.4 GB
Fits Get file →
IQ4_XS0.4 GB 1.8 GB
Fits Get file →
IQ4_NL0.4 GB 1.8 GB
Fits Get file →
Q4_K_M0.4 GB 1.8 GB
Fits Get file →
Q5_K_M0.4 GB 1.9 GB
Fits Get file →
Q6_K0.5 GB 1.9 GB
Fits Get file →
Q8_00.6 GB 2.1 GB
Fits Get file →
BF161.2 GB 2.6 GB
Fits Get file →
F161.2 GB 2.6 GB
Fits Get file →

For 8 GB VRAM, use IQ4_XS (0.37 GB) or Q4_K_M (0.4 GB). For 16 GB VRAM, Q5_K_M (0.44 GB) or Q6_K (0.5 GB) are suitable. Higher quality Q8_0 (0.64 GB) or BF16 (1.2 GB) can be used on 24 GB VRAM or more.

What GPU do you need for LightOnOCR 2 1B Ocr Soup?

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 BF16 (1.2 GB file, about 2.6 GB in use).

Which LightOnOCR 2 1B Ocr Soup quant should you choose?

Q4_K_M is the usual starting point for LightOnOCR 2 1B Ocr Soup: it balances file size and quality well. Here it is a 0.4 GB file.

The Q8_0 file is 0.6 GB, about 1.6 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 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

Architectureqwen3
Parameters596M
Layers28
Trained context16,384 tokens
KV cache (fp16)112 KB per token

Source

Quantized bynoctrex
Licenseapache-2.0
Downloads (HF)2,475
Last updated2026-01-24

Frequently asked questions about LightOnOCR 2 1B Ocr Soup

How much VRAM does LightOnOCR 2 1B Ocr Soup need?

About 1.8 GB at Q4_K_M with an 8K context window. That is the 0.4 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 LightOnOCR 2 1B Ocr Soup on an 8 GB GPU?

Yes, with BF16: the 1.2 GB file needs about 2.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of LightOnOCR 2 1B Ocr Soup GGUF?

It depends on the quant: file is 0.0 GB, IQ4_XS is 0.4 GB, IQ4_NL is 0.4 GB, Q4_K_M is 0.4 GB, Q5_K_M is 0.4 GB, Q6_K is 0.5 GB and 3 more.

What context length does LightOnOCR 2 1B Ocr Soup support?

The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 0.9 GB at 8K and 1.9 GB at 16K.

Where can I download LightOnOCR 2 1B Ocr Soup GGUF?

The files are on Hugging Face in the noctrex/LightOnOCR-2-1B-ocr-soup-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 LightOnOCR 2 1B Ocr Soup 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 LightOnOCR‑2 on 8 GB VRAM?

Yes, using IQ4_XS or Q4_K_M quantizations keeps VRAM usage below 2 GB at 8 k context.

Does the model support multilingual OCR?

The model is trained on diverse data but no explicit language list is provided; it handles multiple languages in practice.

What is the fastest quantization?

IQ4_XS is the smallest file and uses the least memory, but Q8_0 offers higher quality at a modest memory increase.

Similar models

Browse all →

More from noctrex

All by noctrex →

More Lightonocr models

All Lightonocr →

Other tiny models

All tiny models →