Generation, voice and likeness each go to whoever is best at them.
Generation goes to Seedance 2.5. Voice goes to ElevenLabs. Likeness goes to HeyGen. We did not build worse versions of any of them in-house. What we built is the part nobody else does well: one console where the keys, the consent, the costs and the credits all live together.
Text to video and image to video, plus a genuine edit pass on a clip you already have with the source attached automatically. Renders land on the timeline as media with the model, prompt, job id and the real cost stamped on them.
Harvest training audio out of a take you already have, clone, then type corrections into the transcript and hear them in your own voice. Dubbing into other languages rides the same word timings.
- Consent statement recorded in ShowShot before any upload
- Multiple voices per brand, not one per account
- Same word-timing shape the transcript editor already uses
A guided minute on camera inside the recorder trains a look you can send scripts to forever. Photo avatars from a single still for the days you will not sit down for a take.
- Training take captured in the app, not uploaded from elsewhere
- Several looks per brand, switchable per project
- Overlay a twin onto any recording layout
Keys, credits, and who pays for the call.
An account admin decides per workspace whether people bring their own keys, spend the company balance, or both. Keys are stored server side and never handed to a browser. Every job writes back what it truly cost.
Pay providers directly, at cost.
Paste an OpenRouter, ElevenLabs or HeyGen key once. No password prompt every session, no key in local storage, no ShowShot margin on the call. You see exactly the same number the provider bills you.
One balance, every provider.
Plus includes $5 of AI credit a month and Pro includes $25, on ShowShot's keys. Add more to the balance and stop maintaining three vendor accounts: auto top-up at a threshold you set, per-workspace spend caps, and a ledger that shows which project spent what.
Consent is a first-class object. Before a voice or a likeness leaves the building, ShowShot collects and stores the consent statement itself, hosted first, with the provider's approval step after. It is not a checkbox on somebody else's page that you never see again.
Record it once. Deliver it a thousand ways.
No card, no trial countdown, no seat to justify. Record something in the next ninety seconds and see what the link does.