A mobile AI coding IDE that pays token costs with short videos
This mobile-first IDE prototype tries to turn AI coding wait time into a way to pay for model use. Large requests sent to models such as or Fable 5 can take 1 to 3 minutes to process, and that pause can push people to leave the coding screen and lose focus. The workspace includes a terminal and a model switcher.
After a large prompt is sent, a vertical short-video feed appears inside the coding workspace while the model prepares the code. The videos are limited to developer-focused content, such as syntax tips, cloud tools, and developer platforms, instead of random entertainment. The plan is to use ad revenue from those views to cover OpenRouter token bills, making premium AI feel free to the coder.
When the code is ready, the video feed disappears and the workspace returns to normal.
Key points
- The prototype is a mobile-first IDE for on a phone.
- It assumes large prompts can leave the coder waiting 1 to 3 minutes.
- A developer-focused short-video feed appears while the AI model is processing.
- Ad revenue from the videos is meant to pay OpenRouter .
- The feed goes away once the generated code is ready.