PA

PaddleOCR-VL-1.6

PaddleocrGGUF by MungertUpdated 2026-06-06
Download on Hugging Face
Parameters467M
Context128K
Layers18
Architecturepaddleocr
Smallest quant0.0 GB
Quick answer

PaddleOCR-VL-1.6 needs about 1.0 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 20 quants, from 0.0 GB to 0.9 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 0.6 GB.

Compact VLM for robust document parsing

PaddleOCR-VL-1.6 is a 0.467 B parameter vision‑language model built on PaddleOCR‑VL‑1.5. It achieves 96.33 % on OmniDocBench v1.6 and supports document parsing tasks such as text spotting, table recognition, and seal detection.

Good at

  • High accuracy on OmniDocBench v1.6
  • Robust text spotting and table parsing
  • Versatile document recognition tasks

Limits

  • Vision‑only, no text‑only inference
  • Requires 8 k context for memory estimates
  • License: Apache‑2.0

Best for

  • Document OCR in mobile apps
  • Batch processing of scanned forms
  • Automated table extraction

Who it’s for. Developers building document‑centric applications who need a lightweight VLM. The model is suitable for edge devices with limited VRAM and for integration into existing OCR pipelines.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—0.0 GB 0.7 GB
Fits Get file →
IQ4_NL0.3 GB 0.9 GB
Fits Get file →
Q3_K_S0.3 GB 0.9 GB
Fits Get file →
IQ3_XXS0.3 GB 0.9 GB
Fits Get file →
IQ3_XS0.3 GB 0.9 GB
Fits Get file →
IQ3_M0.3 GB 0.9 GB
Fits Get file →
Q3_K_M0.3 GB 1.0 GB
Fits Get file →
Q4_10.3 GB 1.0 GB
Fits Get file →
IQ4_XS0.3 GB 1.0 GB
Fits Get file →
Q4_K_S0.3 GB 1.0 GB
Fits Get file →
Q4_00.4 GB 1.0 GB
Fits Get file →
Q4_K_M0.4 GB 1.0 GB
Fits Get file →
Q5_00.4 GB 1.1 GB
Fits Get file →
Q5_K_M0.4 GB 1.1 GB
Fits Get file →
Q5_10.4 GB 1.1 GB
Fits Get file →
Q6_K_M0.4 GB 1.1 GB
Fits Get file →
Q8_00.5 GB 1.1 GB
Fits Get file →
BF160.8 GB 1.4 GB
Fits Get file →
F160.8 GB 1.4 GB
Fits Get file →
BF160.9 GB 1.6 GB
Fits Get file →

For 8 GB VRAM, use Q4_0 (0.37 GB file, ~1 GB memory at 8 k context). For 4 GB VRAM, use Q3_K_S (0.27 GB file, ~0.9 GB memory). For 2 GB VRAM, use IQ3_XXS (0.27 GB file, ~0.9 GB memory).

What GPU do you need for PaddleOCR-VL-1.6?

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 (0.9 GB file, about 1.6 GB in use).

Which PaddleOCR-VL-1.6 quant should you choose?

Q4_K_M is the usual starting point for PaddleOCR-VL-1.6: it balances file size and quality well. Here it is a 0.4 GB file.

The Q8_0 file is 0.5 GB, about 1.3 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

Architecturepaddleocr
Parameters467M
Layers18
Trained context131,072 tokens
KV cache (fp16)18 KB per token

Source

Quantized byMungert
FamilyPaddleocr
Licenseapache-2.0
Downloads (HF)4,272
Last updated2026-06-06

Frequently asked questions about PaddleOCR-VL-1.6

How much VRAM does PaddleOCR-VL-1.6 need?

About 1.0 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 PaddleOCR-VL-1.6 on an 8 GB GPU?

Yes, with BF16: the 0.9 GB file needs about 1.6 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of PaddleOCR-VL-1.6 GGUF?

It depends on the quant: file is 0.0 GB, IQ4_NL is 0.3 GB, Q3_K_S is 0.3 GB, IQ3_XXS is 0.3 GB, IQ3_XS is 0.3 GB, IQ3_M is 0.3 GB and 14 more.

What context length does PaddleOCR-VL-1.6 support?

The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.6 GB at 32K.

Where can I download PaddleOCR-VL-1.6 GGUF?

The files are on Hugging Face in the Mungert/PaddleOCR-VL-1.6-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 PaddleOCR-VL-1.6 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 PaddleOCR‑VL‑1.6 on 8 GB VRAM?

Yes, the Q4_0 quant requires about 1 GB VRAM at 8 k context, which fits within 8 GB.

Does the model support text‑only inference?

No, PaddleOCR‑VL‑1.6 is vision‑only and requires image input.

What license applies to this model?

The model is released under the Apache‑2.0 license.

Other GGUFs of PaddleOCR-VL-1.5

Similar models

Browse all →

More from Mungert

All by Mungert →

More Paddleocr models

All Paddleocr →

Other tiny models

All tiny models →