What are the hardware requirements for Falcon OCR?

Falcon OCR requires approximately 1.1 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Falcon OCR and how large is it?

Falcon OCR is developed by tiiuae, a model with 269.9m 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 Falcon OCR?

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

Falcon OCR by tiiuae — VRAM 1.1GB

Name: Falcon OCR
Author: tiiuae

Falcon OCR is an open-source language model by tiiuae. Features: 269.9m LLM, VRAM: 1.1GB, License: apache-2.0, LLM Explorer Score: 0.28.

Arxiv:2603.27365 Custom code Document-understanding Eval-results Falcon Falcon ocr Image-to-text Ocr Region:us Safetensors Vision Vision-language

Model Card on HF 🤗: https://huggingface.co/tiiuae/Falcon-OCR

Falcon OCR Parameters and Internals

LLM Name	Falcon OCR
Repository 🤗	https://huggingface.co/tiiuae/Falcon-OCR
Model Size	269.9m
Required VRAM	1.1 GB
Updated	2026-08-08
Maintainer	tiiuae
Model Type	falcon_ocr
Model Files	1.1 GB
Model Architecture	FalconOCRForCausalLM
License	apache-2.0
Tokenizer Class	TokenizersBackend
Vocabulary Size	65536
Torch Data Type	float32

Rank the Falcon OCR 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