What are the hardware requirements for Nanonets OCR2 3B?

Nanonets OCR2 3B requires approximately 7.5 GB of VRAM and supports a context window of 125K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Nanonets OCR2 3B and how large is it?

Nanonets OCR2 3B is developed by nanonets, a model with 3b 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 Nanonets OCR2 3B?

Nanonets OCR2 3B is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Nanonets OCR2 3B by nanonets — VRAM 7.5GB, 125K context

Name: Nanonets OCR2 3B
Author: nanonets

Nanonets OCR2 3B is an open-source language model by nanonets. Features: 3b LLM, VRAM: 7.5GB, Context: 125K, Instruction-Based, LLM Explorer Score: 0.35.

Base model:finetune:qwen/qwen2... Base model:qwen/qwen2.5-vl-3b-... Conversational Endpoints compatible Eval-results Image-text-to-text Image-to-text Instruct Multilingual Ocr Pdf2markdown Qwen2 5 vl Region:us Safetensors Sharded Tensorflow Vqa

Model Card on HF 🤗: https://huggingface.co/nanonets/Nanonets-OCR2-3B

Nanonets OCR2 3B Parameters and Internals

LLM Name	Nanonets OCR2 3B
Repository 🤗	https://huggingface.co/nanonets/Nanonets-OCR2-3B
Base Model(s)	Qwen/Qwen2.5-VL-3B-Instruct Qwen/Qwen2.5-VL-3B-Instruct
Model Size	3b
Required VRAM	7.5 GB
Updated	2026-08-08
Maintainer	nanonets
Model Type	qwen2_5_vl
Instruction-Based	Yes
Model Files	5.0 GB: 1-of-2 2.5 GB: 2-of-2
Model Architecture	Qwen2_5_VLForConditionalGeneration
Context Length	128000
Model Max Length	128000
Transformers Version	4.55.4
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	151936
Torch Data Type	bfloat16
Errors	replace

Rank the Nanonets OCR2 3B 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 Nanonets OCR2 3B

Best Alternatives	Context / RAM	Downloads	Likes
Qwen2.5 VL 3B Instruct	125K / 7.5 GB	2710	6