← Library · Frontier
Perceptron Introduces Mk1 for Video Understanding and Embodied Reasoning
Perceptron has released Mk1, a new closed-source model specifically designed for video understanding and embodied reasoning. Mk1 analyzes video at a dynamic frame rate, up to 2 FPS, across a 32K-token context window, and can pinpoint specific moments within long streams with structured timecodes. It achieves frontier performance in image, video, and embodied reasoning, matching models like the Gemini Pro series, at a significantly lower cost.
Why it matters
Mk1 makes advanced video understanding and embodied reasoning deployable and cost-effective, paving the way for more sophisticated physical AI agents that can see, reason, and act in the physical world.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free