What are the hardware requirements for OvisOCR2?

OvisOCR2 requires approximately 1.7 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed OvisOCR2 and how large is it?

OvisOCR2 is developed by ATH-MaaS, 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 OvisOCR2?

OvisOCR2 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

OvisOCR2 by ATH-MaaS — VRAM 1.7GB

Name: OvisOCR2
Author: ATH-MaaS

OvisOCR2 is an open-source language model by ATH-MaaS. Features: 0.8b LLM, VRAM: 1.7GB, License: apache-2.0, LLM Explorer Score: 0.47.

Arxiv:2607.13639 Base model:finetune:qwen/qwen3... Base model:qwen/qwen3.5-0.8b Conversational Document-parsing Endpoints compatible Formulas Image-text-to-text Markdown Multimodal Ocr Qwen3 5 Region:us Safetensors Tables Vllm

Model Card on HF 🤗: https://huggingface.co/ATH-MaaS/OvisOCR2

OvisOCR2 Parameters and Internals

LLM Name	OvisOCR2
Repository 🤗	https://huggingface.co/ATH-MaaS/OvisOCR2
Base Model(s)	Qwen/Qwen3.5-0.8B Qwen/Qwen3.5-0.8B
Model Size	0.8b
Required VRAM	1.7 GB
Updated	2026-08-08
Maintainer	ATH-MaaS
Model Type	qwen3_5
Model Files	1.7 GB
Model Architecture	Qwen3_5ForConditionalGeneration
License	apache-2.0
Model Max Length	262144
Transformers Version	4.57.0.dev0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Errors	replace

Rank the OvisOCR2 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 OvisOCR2

Best Alternatives	Context / RAM	Downloads	Likes
...e 4.6 Opus Reasoning Distilled	0K / 1.8 GB	817	10
GRM 2.5 Air	0K / 1.8 GB	27	8
Qwen3.5 0.8B MLX 4bit	0K / 0.6 GB	5113	5
Qwen3.5 0.8B OptiQ 4bit	0K / 0.2 GB	1494	22