Introducing Voice Mode

Open the canvas on Krea for iPad and tap the microphone icon. Start drawing, and start talking. You might say "make the background warmer" or "add more contrast to the left side" while your pencil is still on the screen. The AI listens, interprets what you're asking in the context of what you're drawing, and applies the change live.

There's no separate prompt bar and no submit button. You don't compose instructions and send them off. You just describe what you want as it occurs to you, the same way you'd talk to a collaborator sitting next to you. The model picks up on your intent and adjusts the output while you keep working.

Because the response is continuous, you can course-correct as you go. Say "darker" and watch it shift. Say "not that much" and it pulls back. The feedback loop tightens to the point where speaking and drawing feel like a single gesture.

Voice Mode ships first on iPad because that's where it makes the most sense. A tablet is already the closest thing to working on paper. You hold it, you draw on it directly, you rotate it. Adding voice to that feels obvious in a way it wouldn't on a desktop, where your hands are already on a keyboard.

Touch and voice together map to something people already know how to do. You point at a region with your finger and say what you want changed. You sketch a rough shape and describe the material. The inputs are natural, not learned. There's no interface vocabulary to memorize, no modifier keys or nested menus. You just work.

This is the direction we keep pushing toward at Krea. The most capable tools should be the simplest to use. Voice Mode is available now in Krea for iPad.