---
license: apache-2.0
pipeline_tag: image-text-to-text
language:
- en
- fr
- de
- es
- it
- nl
- pt
- sv
- da
- zh
- ja
library_name: transformers
tags:
- ocr
- document-understanding
- vision-language
- pdf
- tables
- forms
- llama.cpp
base_model:
- lightonai/LightOnOCR-2-1B
---

# **LightOnOCR-2-1B-f32-GGUF**

> LightOnOCR-2-1B from lightonai is LightOn's flagship 1B-parameter end-to-end vision-language OCR model—the recommended variant for most tasks—refined via RLVR training on a 2.5x scaled 43M-page corpus with enhanced French, arXiv, scan, and LaTeX coverage for converting PDFs, scans, and document images into clean, naturally ordered text at 3.3× Chandra OCR speed, 1.7× OlmOCR, 5× dots.ocr, and 5.71 pages/s on H100 (~<$0.01/1k pages) while achieving state-of-the-art 83.2±0.9 on OlmOCR-Bench (outperforming Chandra-9B by 1.5+ points) across tables, receipts, forms, multi-column layouts, and math without brittle pipelines. Part of the fully differentiable LightOnOCR-2 family (including bbox variants for image localization, base models for fine-tuning, and soup merges), it uses a native-resolution ViT encoder (from Mistral-Small-3.1), MLP projector, and Qwen3 decoder with 1540px longest-edge preprocessing (200 DPI PDFs) for superior accuracy on degraded scans, scientific docs, and European languages under Apache 2.0, supporting LoRA/PEFT fine-tuning via Transformers for domain adaptation.

## LightOnOCR-2-1B [GGUF]

| File Name | Quant Type | File Size | File Link |
| - | - | - | - |
| LightOnOCR-2-1B-BF16.gguf | BF16 | 1.2 GB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B-BF16.gguf) |
| LightOnOCR-2-1B-F16.gguf | F16 | 1.2 GB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B-F16.gguf) |
| LightOnOCR-2-1B-F32.gguf | F32 | 2.39 GB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B-F32.gguf) |
| LightOnOCR-2-1B-Q8_0.gguf | Q8_0 | 639 MB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B-Q8_0.gguf) |
| LightOnOCR-2-1B.mmproj-BF16.gguf | mmproj-BF16 | 829 MB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B.mmproj-BF16.gguf) |
| LightOnOCR-2-1B.mmproj-F16.gguf | mmproj-F16 | 819 MB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B.mmproj-F16.gguf) |
| LightOnOCR-2-1B.mmproj-F32.gguf | mmproj-F32 | 1.64 GB | [Download](https://huggingface.co/prithivMLmods/LightOnOCR-2-1B-f32-GGUF/blob/main/LightOnOCR-2-1B.mmproj-F32.gguf) |

## Quants Usage 

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant
types (lower is better):

![image.png](https://www.nethype.de/huggingface_embed/quantpplgraph.png)