Unlimited-OCR — MLX bf16

This repository contains the unquantized bfloat16 MLX conversion of Baidu's official baidu/Unlimited-OCR checkpoint. It is published for Apple Silicon inference with mlx-vlm and maintained for use by shuuul/aimd.

The model architecture, tokenizer, processor configuration, prompts, and MIT license come from the official checkpoint. Refer to the official model card and Unlimited-OCR repository for model details, capabilities, limitations, and attribution.

Conversion details

Weights are stored as bfloat16 without weight quantization.

Conversion command:

mlx_vlm.convert \
  --hf-path baidu/Unlimited-OCR \
  --mlx-path ./Unlimited-OCR-bf16 \
  --dtype bfloat16

Use with AIMD

On macOS/Apple Silicon, install AIMD and run OCR directly:

uv tool install --force "aimd-tool @ git+https://github.com/shuuul/aimd.git@main"
aimd scan.pdf --task ocr --model unlimited_ocr_bf16

unlimited_ocr defaults to the 4-bit repository. AIMD runs each page in Unlimited-OCR's single-image gundam mode with prompt document parsing., cropping=True, base_size=1024, image_size=640, and a sliding-window no-repeat n-gram guard. See the AIMD README for the complete runtime behavior.

Use with mlx-vlm

pip install -U "mlx-vlm>=0.6.8"
python -m mlx_vlm.generate \
  --model mlx-community/Unlimited-OCR-bf16 \
  --image scan.png \
  --prompt "document parsing." \
  --max-tokens 8192 \
  --temperature 0

Unlimited-OCR support first shipped in mlx-vlm 0.6.4; 0.6.8 or newer is recommended to match the conversion environment.

Available MLX weights

License and attribution

The upstream checkpoint is released under the MIT license. This repository is a format conversion and does not modify the upstream model architecture or claim ownership of Baidu's model.