Raspberry Pi AI Camera Module 12.3 MP Sony’s IMX500 Intelligent Vision Sensor - Official

Raspberry Pi AI Camera Module – Key Specifications

Sensor & AI:

• Sony IMX500 – 12.3MP (4056×3040)
• Built-in AI accelerator for running ML models on-camera
• Supports TensorFlow, PyTorch (int8/uint8 models)
• Onboard RP2040 microcontroller

Image & Video Output:

• Max resolution: 4056×3040 @ 10fps
• 2x2 Binned mode: 2028×1520 @ 30fps
• Output formats: RAW10, YUV/RGB, Region of Interest (ROI), tensor metadata

Optics:

• Focal length: 4.74mm, Aperture: f/1.79
• Field of View (FOV): 78.3° diagonal
• Manual focus (20cm to ∞)
• IR-cut filter included (not IR-sensitive)

AI Capabilities:

• Input tensor size: up to 640×640 pixels
• Onboard model storage: ~8.4MB
• Inference directly on sensor (no need for external accelerator)

Physical Specs:

• Dimensions: 25 × 24 × 11.9 mm
• Sensor size: 7.857 mm (1/2.3” type)
• Pixel size: 1.55 µm x 1.55 µm

Connectivity & Compatibility: 

• Compatible with Raspberry Pi Zero 2, 3, 4, 5
• PCB size and mount matches Camera Module 3

Other Info:

• Operating Temp: 0°C to 50°C
• Production supported until at least January 2028

Datasheet

• Official product page​