A 6-bit, group-size-64 MLX quantization of PaddlePaddle/PaddleOCR-VL-1.6, for fast on-device OCR on Apple Silicon. The architecture is unchanged — this repo only re-quantizes the original weights.

This is the size/quality sweet spot: in side-by-side OCR benchmarking it matches the 8-bit variant's quality, and — unlike 4-bit — it does not slip into repetition / hallucination loops on hard scripts, at a smaller footprint than 8-bit.

Other tiers: 8-bit (slightly larger, same quality) and 4-bit (smallest, but prone to loops on hard/rare scripts).

Apache-2.0, inherited from the base model. All credit for the model goes to the PaddlePaddle / PaddleOCR team — this repository only provides an MLX-quantized copy.