Skip to content
View original post on X: OpenBMBOfficial
Published
28/100

MiniCPM-OCR models now run on Apple Silicon via MLX

AIOpenBMB says its MiniCPM-OCR 1.0 (1.3B) and MiniCPM-OCR Jev (2B) now run on Apple Silicon, in a partnership with the Neywa Labs team. On an M5 Max, OCR 1.0 reaches 6,339 tok/s prefill and 220–235 tok/s decode, while Jev's median OCR latency is 62.9 ms.
OpenBMBVerified on XOfficial
@OpenBMB
View on X

MiniCPM-OCR 1.0 (1.3B) and MiniCPM-OCR Jev (2B) now run on Apple Silicon, thanks to our partnership with the @neywalabs team! 🚀

On an M5 Max, OCR 1.0 reaches 6,339 tok/s prefill and 220–235 tok/s decode, while Jev scores OCR results in just 62.9 ms median latency.

Check out the blog: https://runnativ.ai/blog/minicpm-ocr-on-mlx-vlm

Source details

OpenBMB · x.comPublished · added here

Original title: MiniCPM-OCR 1.0 (1.3B) and MiniCPM-OCR Jev (2B) now run on Apple Silicon, thanks to our partnership with the @neywalabs team! 🚀

Topics

On-device AIOpen-source ecosystemDeployment & engineeringProduct updates