AI Storymaker

Create three characters and an optional situation, then generate a short story with OpenAI, Anthropic's Claude, and Google Gemini each playing a character.

How it works

1. You create and name three characters, either from a variety of presets or by entering your own values.
2. A random situation is suggested. You can change it, ask for a different one, or delete it altogether.
3. The characters and situation are sent to an AWS Lambda function, which writes the story using models from OpenAI, Anthropic and Google, each playing one of the characters in a back-and-forth skit.
4. After a moment, the story prints out in the results box, and a button appears that generates an audio file of a narrator reading it. The story is usually too long to hand to the speech function directly, and narrating it takes longer than API Gateway will wait, so the text is saved to S3 and the audio is generated from there by OpenAI's audio model.
5. Audio can take up to a couple of minutes. The page checks every 5 seconds for the finished file, and an audio player appears as soon as it is ready.