The Raspberry Pi AI Camera is a compact camera module that combines a Sony IMX500 sensor with a 12-megapixel image sensor and a built-in AI accelerator. This enables the development of advanced computer vision applications without a separate computing unit.
The camera automatically adds tensor metadata to captured photos and videos, freeing the Raspberry Pi’s own processor for other tasks. Software support through the libcamera, Picamera2 and rpicam-apps libraries makes the camera easy to use for beginners while offering extensive customizability for more experienced users.

Features

• Sensor: Sony IMX500 Intelligent Vision Sensor
• Integrated neural network inference accelerator
• Resolution: 12.3 megapixels
• Sensor size: 6,287 × 4,712 mm (1/2,3")
• Resolution (horizontal/vertical): 4056 × 3040 pixels
• Video modes: 2028 × 1520p @ 30 Hz, 4056 × 3040p @ 10 Hz
• Focus: adjustable, sharpness range approx. 20 cm – ∞
• Field of view: 66 ± 3° (horizontal), 52.3 ± 3° (vertical)
• Tensor metadata support: libcamera, Picamera2 and rpicam-apps
• Compatibility: all Raspberry Pi computers
• Cables: FPC15 (CSI) 200 mm (Pi 3/4) and FPC15–FPC22 (CSI) 200 mm (Pi 5/Zero/Zero 2)
• Dimensions: 25 × 24 × 11.9 mm
• Weight: 6 g