AI UGC Video Generator

Playad lets you produce UGC-style video ads featuring an on-screen presenter with lip-synced dialogue and a matched voice — without sourcing real creators. Start a conversation in Chat, describe your product and audience, attach a presenter image or choose from the avatar library, and Playad builds a production flow. You review the proposed flow before any generation runs, then refine performance, voice, and copy in Flow Editor. Review the generated presenter-led video before downloading or using it in an ad.

What you can do

On-screen presenter with lip-synced speech

Bind a visible presenter to exact dialogue and a matched voice. The Seedance presenter speech workflow aligns audible delivery with the on-screen performance so speech and lip movement correspond.

Avatar library and custom avatar creation

Browse saved presenters in the Avatars library or create one from scratch by generating an appearance from a prompt or uploading your own image, then setting a name, profile, and voice before generating the avatar.

Voice selection and accent matching

Playad searches available voices by language, speaker gender, and requested accent or qualities, then verifies the returned voice matches the resolved speaker before binding it to the scene.

Chat-based brief entry

Describe the product, audience, message, desired duration, and placement in Chat. Attach product images or a presenter reference. Chat proposes a flow you can review and adjust before generation begins.

Flow review and editing before generation

Open the flow card in Chat to review the Flow Preview. Continue in Flow Editor on desktop to adjust scene timing, dialogue copy, presenter performance notes, and composition before committing to generation.

Saved work in Projects and Creations

Generated videos appear in the originating conversation and in Creations for the same brand. Use Projects to reopen an associated editable project and continue refining the flow.

How it works

1. Open Chat and describe your ad

Select your team and brand, then choose New Chat. Describe the video: product, target audience, key message, desired duration, and placement size. Attach a product image or presenter reference if you have one.

2. Choose or create a presenter

Reference an avatar from your Library in the brief, or open Avatars to browse by gender, age, and ethnicity. To create one, choose Create from Scratch, generate or upload an appearance, set the profile, select a voice, and generate the avatar. Confirm the mouth is clearly visible before saving.

3. Review the proposed flow

When Chat proposes a flow, open its flow card and review the Flow Preview. Check the presenter binding, scene count, dialogue copy, and voice assignment. Ask Chat to adjust the concept or any step before continuing.

4. Refine performance and dialogue in Flow Editor

Open the flow in Flow Editor on desktop. Adjust the exact dialogue words, delivery notes, gaze and gesture direction, scene timing, and any product-only inserts. Keep the presenter identity consistent across scenes.

5. Generate and review the video

Run generation from the current flow's controls. Creating a flow is not the same as completing generation — follow the review and generation controls shown in the editor. Check lip synchronisation against the audible speech in the output.

6. Find and use the result

Locate the generated video in the conversation and in Creations for the same brand. Download it or use available actions. Reopen the project in Projects if further edits are needed.

Using Playad: product screens

Example request

Create a 30-second vertical UGC video for my protein bar brand. The presenter should speak directly to camera and say: 'I've tried every bar out there — this one actually tastes good and keeps me full.' Use a female presenter in her late 20s with an upbeat tone. Attach product image.

Hypothetical example: skincare product launch

Illustrative example — not a product screenshot or customer result

Input: A user opens Chat, selects their skincare brand, and types: 'Create a 30-second vertical UGC-style video for our new SPF serum, aimed at women 25–40. Use the avatar I saved last week. She should look at camera and say: This is the only SPF I've worn every day for a month — and I mean every day. Highlight the lightweight texture.' They attach a product flat-lay image.

Review: Chat proposes a single-scene flow with the saved avatar as presenter, the exact supplied dialogue bound to her performance, and a searched voice matched to her profile. The user opens the flow card, confirms the Flow Preview shows the correct presenter and copy, then opens Flow Editor to adjust her gesture direction during the word 'lightweight'. They run generation and review lip sync against the audio in the Creations output.

Before you start

On-screen speech requires a visible presenter with an unobstructed mouth; a hidden or cropped mouth must be corrected before generation.

Reference audio used for voice guidance must fall within the selected Seedance model's per-reference and combined duration limits; longer speech must be split at natural boundaries.

An on-screen speaking presenter and off-screen narration use different workflows. Describe which you need in the video brief.

Custom avatar creation depends on the team's plan limit; reaching the limit opens the upgrade flow rather than allowing another creation.

Baked subtitles and captions are not added during video generation; captions are a separate supported Flow step.

Flow Editor is available on desktop; mobile users can review results in Chat and Creations but must continue detailed editing on a computer.

Frequently asked questions

Do I need a real person or footage to make a UGC-style video?

No. Playad generates an on-screen presenter from an avatar you create or select from the library. You can generate an appearance from a prompt or upload an image, then assign a profile and voice without sourcing real-creator footage.

How does Playad match the voice to the presenter?

Playad searches available voices using the known language, speaker gender, and any requested accent or quality you specify. It verifies the returned voice matches the resolved speaker before binding it to the scene. For English, an accent can be inferred from the specified market if none is stated explicitly.

Can I supply my own audio for the presenter's speech?

Yes, via the reference-audio route. The audio must fall within the selected Seedance model's per-reference and combined duration limits. Playad verifies actual media duration; speech that exceeds the limit must be split at natural boundaries. The reference guides a new performance rather than copying the exact waveform.

What is the difference between creating a flow in Chat and completing generation?

Chat proposes a flow you can review in the Flow Preview before anything is generated. Creating the flow is a planning step. You follow the current flow's review and generation controls — in Chat or Flow Editor — to actually run generation and produce the video.

Where do I find the finished video after generation?

Generated videos appear in the originating Chat conversation and in Creations for the same brand. If an associated editable project exists, you can reopen it in Projects to make further changes.

Can I add captions to the UGC video?

Captions are not baked into the video during generation. They are a separate supported step available within the Flow, distinct from the presenter speech and lip-sync process.

Product guides