What are the hardware requirements for LightOnOCR 3 0.8B MTP Repacked?

LightOnOCR 3 0.8B MTP Repacked requires approximately 1.8 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed LightOnOCR 3 0.8B MTP Repacked and how large is it?

LightOnOCR 3 0.8B MTP Repacked is developed by prithivMLmods, a model with 0.8b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate LightOnOCR 3 0.8B MTP Repacked?

LightOnOCR 3 0.8B MTP Repacked is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

LightOnOCR 3 0.8B MTP Repacked by prithivMLmods — VRAM 1.8GB

Name: LightOnOCR 3 0.8B MTP Repacked
Author: prithivMLmods

LightOnOCR 3 0.8B MTP Repacked is an open-source language model by prithivMLmods. Features: 0.8b LLM, VRAM: 1.8GB, License: apache-2.0.

Base model:lightonai/lightonoc... Base model:quantized:lightonai... Conversational Document-understanding En Endpoints compatible Eval-results Forms Gguf Image-text-to-text Llama-cpp Mtp Ocr Pdf Qwen3 5 Region:us Safetensors Sharded Tables Tensorflow Vision Vision-language Visual-grounding

Model Card on HF 🤗: https://huggingface.co/prithivMLmods/LightOnOCR-3-0.8B-MTP-Repacked

LightOnOCR 3 0.8B MTP Repacked Parameters and Internals

LLM Name	LightOnOCR 3 0.8B MTP Repacked
Repository 🤗	https://huggingface.co/prithivMLmods/LightOnOCR-3-0.8B-MTP-Repacked
Base Model(s)	lightonai/LightOnOCR-3-0.8B lightonai/LightOnOCR-3-0.8B
Model Size	0.8b
Required VRAM	1.8 GB
Updated	2026-11-06
Maintainer	prithivMLmods
Model Type	qwen3_5
Model Files	0.7 GB: 1-of-3 0.7 GB: 2-of-3 0.4 GB: 3-of-3
Supported Languages	en
Model Architecture	Qwen3_5ForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	5.5.4
Tokenizer Class	TokenizersBackend
Padding Token	<\|endoftext\|>
Errors	replace

Rank the LightOnOCR 3 0.8B MTP Repacked capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to LightOnOCR 3 0.8B MTP Repacked

Best Alternatives	Context / RAM	Downloads	Likes
... 0.8B Weight Space Programming	0K / 1.8 GB	1297	90
OvisOCR3	0K / 1.7 GB	4972	109
... 0.8B Weight Space Programming	0K / 1.8 GB	227	56
Xiaomi OCR 0	0K / 1.8 GB	16009	164
Qwen3.5 0.8B	0K / 1.8 GB	2839884	655
OvisOCR2	0K / 1.7 GB	83978	506
MinerU2.5 Flash 0.8B	0K / 1.7 GB	23	1
Qwen3.5 0.8B Astra WeightEdit	0K / 1.8 GB	141	19
LightOnOCR 3 0.8B	0K / 1.7 GB	1318	48
Qwen3.5 0.8B Base	0K / 1.8 GB	828054	92

Note: green Score (e.g. "73.2") means that the model is better than prithivMLmods/LightOnOCR-3-0.8B-MTP-Repacked.