Your operator writes a reply as usual. We turn it into a 6-second clip: the persona moves, speaks the line and says the user's first name, in many languages. You sell the clip as a paid video message.
The text reply arrives at once. The chat goes on as normal.
A bubble announces the video: "Lea is recording a video for you…", with a progress ring.
The bubble turns into a blurred thumbnail with its price in your credits.
On unlock, the clip plays with sound. She says his first name and the line. The clip carries an AI label.
Each unlocked clip joins his "Her videos" gallery. The collection brings the user back.
Your team and your workflow stay the same. Your operators keep writing the replies. A "Send as video" button next to each message sends that reply to us. The persona's story and the conversation stay with you. We receive only what the clip needs: the persona, the scene, the pose, the line, the first name and the language.
POST /v1/replies
{
"persona": "lea-01",
"scene": "sofa-evening",
"pose": "P063",
"lang": "en",
"first_name": "Tom",
"line": "Of course I'm thinking of you… come and see me."
}
→ webhook in under 40 s: { "status": "accepted", "clip_url": "…", "ai_mark": true }
Pick the pose in the pose gallery: 80 poses, each seen from 5 angles. Its ID (for example P063) becomes a parameter of the video.
"Send as video" next to each message. One click and the clip goes out.
Personas, scenes, languages, which messages may go out as video, and test results.
On any reply. A paid media item is worth 10 to 15 text messages today.
A welcome video at sign-up. The persona says his first name.
A personal clip when the user stops replying.
A thank-you clip after a tip or a credit purchase.
Everything runs on our side. You send a line and get a clip back. You need no GPUs, no models and no video team.
Datacenter-class GPUs across several cloud providers. The fleet grows with your peak hours and shrinks when they pass. You never pay for idle machines.
Tested end to end. A clip takes about 15 to 25 seconds of GPU time, so it is back in under 40 seconds, with the AI mark already in it.
No setup fee, no hardware, no minimum term. A clip that fails our checks is not billed.
With us, a pilot on one persona starts in days, and your cost follows your sales.
| Plan | Monthly minimum | Clips included | Price per clip | Extra clips |
|---|---|---|---|---|
| Starter | €2,000 | 5,000 | €0.40 | €0.30 |
| Growth | €5,000 | 20,000 | €0.25 | €0.18 |
| Volume | €10,000 | 100,000 | €0.10 | €0.10 |
All prices exclude VAT. Free integration. Many languages supported. The pose gallery is included. Only accepted clips are billed. Every plan includes an SLA: 95% of clips delivered in under 40 seconds.
Each clip is built to order. You choose:
The three clips below are examples of that process, from suggestive to explicit.
Adults only. These demos show nudity and an explicit sexual scene. Fully synthetic characters, generated by AI.