Introduction
The Someone public API lets you create AI characters and generate images, video, audio, and cloned voices over HTTPS — with API keys, credits, and durable async jobs.
The Someone API is a REST API for building with AI characters. Create a character once, then generate images, video, lipsync, and speech for it — or clone a voice from an audio sample. You authenticate with an API key, spend credits per generation, and poll a resource until it’s ready.
Quickstart
Create your first character and generate an image in a few curl calls.
Authentication
Create an API key, set the Authorization header, and understand scopes.
How it works
The base URL is https://api.someone.app. Every resource lives under /v1/.
The API validates your request, reserves credits, and hands the work to a separate engine that runs the actual AI generation as a durable job. You get back a resource with a status, and you poll that resource until it’s completed (or ready, or failed).
Your code → api.someone.app → engine → AI providers
↓
cdn.someone.app (your finished media)
Because generation is asynchronous, most create endpoints return 201 immediately with a status of processing — not the finished result. See Async results for the polling loop.
What you can build
Characters
Create a consistent AI character with a headshot and full-body image.
Generations
Generate images, video, and lipsync “talk” clips for a character.
Audio
Turn text into speech using a cloned or preset voice.
Voices
Clone a voice from an audio sample, or design one from a prompt.
Current availability
Before you start
- A Creator or Studio plan. API key creation and every authenticated request require an active plan. Without one, requests return
403 plan_required. - An API key. Created once in the Someone app settings; see Authentication.
- Credits. Every generation spends credits from your balance. See Pricing.