Features
Screenshot Capture
Capture visual context from the preview panel and send it to the AI.
Overview
One of Coodeen's most powerful features is the ability to capture screenshots from the live preview and include them in your conversation with the AI. This gives the AI visual context about what you're seeing.
How It Works
- Click the screenshot button in the preview panel toolbar
- Select an area — An overlay appears, letting you click and drag to select a region
- Attach to message — The screenshot is automatically attached to your next message
- AI sees it — Vision-capable models can analyze the screenshot and respond accordingly
Use Cases
- UI bugs — Show the AI a visual glitch and ask it to fix the CSS or component
- Layout issues — Capture a section that isn't aligned correctly
- Design feedback — Ask the AI to adjust colors, spacing, or typography based on what you see
- Element selection — Select specific UI elements for targeted fixes
Vision Model Requirements
Screenshot analysis requires a vision-capable model:
- OpenAI: GPT-4o, GPT-4.1
- Anthropic: Claude Sonnet 4, Claude Opus 4
- Google: Gemini 2.5 Pro, Gemini 2.5 Flash
Models without vision support (like GPT-4.1-nano) cannot process image inputs.