Opening the Composer
Click Create post on a character, or Open composer on the teams home. It opens on the post type you used last with that character (Social Post the first time).
- Left: the team's cast (switch characters without leaving), the character's face, vibes, voice and pack, and the five post types with their prices.
- Middle: the brief, then, pinned at the bottom, what it runs on, the price, what you will have left, and Generate.
- Right: the result, with the character's recent posts underneath.
The five post types
| Post type | What you get | Credits |
|---|---|---|
| UGC Ad | A talking video pitching your product, or a product-in-hand photo | Photo 9; talking video per second, by engine (see below) |
| Social Post | A photo of them living their life, with a caption | 9 |
| Backstage Clip | A candid motion video, with spoken lines if you write them | Per second, from 24. See Backstage Clips |
| Vlog Entry | A selfie video talking to camera about their day | Per second, by engine (see below) |
| Personal Musing | An opinion post in their voice, with a selfie | 9 |
Writing the caption and script is free; you pay for the media. Talking videos (UGC Ad talking, Vlog Entry) are one take in the character's voice, priced per second by the engine you pick on the form:
| Engine | 720p |
|---|---|
| Grok | 7 cr/s |
| Mini (default) | 12 cr/s |
| Fast | 35 cr/s |
| 2.0 | 58 cr/s |
| 2.5 | 89 cr/s |
The take is as long as the script needs, so a 10-second read on Mini is 120 credits and on Grok is 70. Takes run up to 15 seconds (up to 30 on 2.5). With Exact voice switched on, the lip-sync pass is added on top at its own per-second rate (PixVerse 12, VEED 2), billed on the real length of the voice.
Under the engine buttons, a line tells you which voice the next take uses before you press Generate: your recording (on the Seedance engines), a description of the voice (on Grok, which cannot take a recording), or Exact voice with the lip sync you picked. It updates as you switch engines.
What each asks for
- UGC Ad: Talking video or Product-in-hand photo, what they are selling, a call to action, and an optional product photo (From my gallery or Upload).
- Social Post: a scene idea: clothes, light, place, action. Hair and eye notes are ignored on purpose, because the face is locked.
- Vlog Entry: what the entry is about, or leave it blank and they riff on their day.
- Personal Musing: a topic, or leave it blank and they pick one.
Photo posts also offer:
- References: up to 6 extra images (a location, outfit or product) from your gallery or an upload, used alongside the character's pack.
- Format: Feed 3:4, Story 9:16 or Square 1:1.
Generating
The line above Generate shows the price and what you will have left. If a fixed-price post costs more than you have, Generate is replaced by Top up N credits to generate. When a talking video's length is not settled yet, or Exact voice is on, its price is an estimate, so a short balance warns but does not block it.
While it works:
- The Generate button becomes a progress bar with a clock.
- The right panel shows a large timer, the current step and the step list. The caption appears as soon as it is written, before the media is done.
- Once the final render has been handed off, it says Safe to leave this page now: the post finishes on our side and lands in your gallery. Before that, keep the page open.
If something fails, the panel says what happened and offers Try again and Edit brief. Your brief is kept.
The result
- The caption is editable, with a count against the 2,200-character limit. Copy and Send with Social-Cast use your edited text.
- It is saved to your gallery automatically.
- Download saves the file to your device; Copy copies the caption.
- Send with Social-Cast opens Social-Cast in a new tab with the caption and media loaded; the Composer stays open behind it.
- Make another returns to the start with your brief kept.
- Recent posts under the result show the character's latest posts; click one to see it again. They are kept in this browser; your gallery has everything.