Published on June 17, 2026, the PaddleOCR 3.7 announcement highlights an ONNX Runtime backend and CPU and GPU inference options. It says PP-OCRv6 models can be up to 3.9 times faster on CPU.
On June 17, 2026, an announcement about PaddleOCR 3.7 reported the addition of an ONNX Runtime inference backend. It also cited CPU inference via OpenVINO and GPU inference via CUDA and TensorRT. According to the text, PP-OCRv6 models are up to 3.9 times faster on CPU; the announcement does not detail the conditions for that result.
The update is relevant to engineers comparing backends and hardware for deploying OCR across platforms. To check the details, consult the original announcement and verify its versions, configurations, and benchmark conditions before comparing performance. If you use AI to study or apply the material with organizational documents, remove or protect personal and confidential data before submitting content.