Raspberry Pi AI Camera (12.3 MP, Sony IMX500 Intelligent Vision Sensor)

The Raspberry Pi AI Camera combines a 12.3-megapixel Sony IMX500 sensor with on-board neural processing, executing machine learning models directly on the module. Compatible with all Raspberry Pi computers, it is an ideal choice for makers, students, and engineers building real-time vision systems without extra accelerator hardware.

On-board intelligent vision processing

The integrated Sony IMX500 sensor executes machine learning inferencing directly on the camera module, removing the need for a separate hardware accelerator. It supports tensor input sizes up to 640 x 640 in int8 or uint8 formats with dedicated on-board memory for model weights and processing.

Detailed 12.3 MP imaging and optics

With a 4056 x 3040 resolution and 1.55 µm pixel pitch, the sensor captures crisp visual detail across a 66-degree horizontal field of view. The module features an f/1.79 aperture, an integrated IR cut filter, and manually adjustable focus from 20 cm to infinity.

Native Raspberry Pi software support

Fully compatible with all Raspberry Pi computers, the module connects smoothly to the standard software stack. Native integration with libcamera and Picamera2 simplifies model deployment and video pipeline configuration.

Versatile stream outputs and metadata

The camera delivers Bayer RAW10 and ISP-processed YUV/RGB video streams alongside region-of-interest and tensor metadata. It outputs full-resolution 4056 x 3040 footage at 10 fps or 2x2 binned 2028 x 1520 video at up to 30 fps.

Compact profile for embedded builds

Measuring 25 x 24 x 11.9 mm and supplied with a 200 mm ribbon cable, the board fits easily into tight project enclosures. It is well-suited for autonomous robotics, security monitoring installations, and remote environmental sensing stations.

Downloads & documentation

• Product Brief