How to Make TikTok Videos With AI: 9:16, Captions, Voice

Make TikTok videos with AI: create a vertical 9:16 clip in Symphony, add auto captions and text-to-speech in the app, and turn on the AI label.

Adam Wang
18/10/2026 · 5 min read
How to Make TikTok Videos With AI: 9:16, Captions, Voice
In this article (6)
  1. Vertical 9:16 is the format TikTok recommends
  2. Create a vertical clip with AI: Symphony Creative Studio is built for TikTok
  3. Captions: let TikTok generate them, then fix each line
  4. Voice: use text-to-speech right in the app
  5. Do AI-made TikTok videos need a label?
  6. Four common mistakes when you make TikTok videos with AI

The simplest way to make TikTok videos with AI is to create a vertical 9:16 clip with an AI tool, then add captions and a voice inside the TikTok app. Before you post, turn on the AI-generated content label.

Vertical 9:16 is the format TikTok recommends

TikTok's In-Feed ad specs page, updated June 2026, lists vertical 9:16 as the recommended format, at a minimum of 540x960 px. Horizontal 16:9 and square 1:1 are still accepted.

A hand holding a phone upright, with the TikTok logo on the screen
TikTok viewers hold their phones upright, so a 9:16 clip fills the screen. Photo: Solen Feyissa/Unsplash.

So pick the vertical format when you create the clip. Do not make a horizontal clip and crop it later. Cropping afterward usually cuts off both sides of the frame, and sometimes the subject too. Guides on which aspect ratio to use for each platform are in News.

Create a vertical clip with AI: Symphony Creative Studio is built for TikTok

The tool closest to TikTok is Symphony Creative Studio. TikTok's About page describes it as a free AI video creation tool for both organic content and ads. You sign in with a TikTok for Business account.

TikTok's About Symphony Creative Studio page: a free AI video tool for organic content and ads
The introduction and four benefits of Symphony Creative Studio. Screenshot: TikTok Business Help Center, Oct 9, 2026.

The video generation guide lists three ways to create a video. They are Text to video (from a text description), Image to video (a still image brought to life), and Reference to video (from several sample images, videos, and audio). Each generation usually takes a few minutes. In the editor, you add music, voice-over, captions, text, and stickers before you download.

TikTok advises simple prompts that describe the scene you want to see. A good example is "A person runs on a treadmill in a gym." Avoid long, roundabout prompts or back-and-forth dialogue.

Captions: let TikTok generate them, then fix each line

TikTok has two kinds of captions, according to its accessibility page. The first is auto-generated captions, made with speech recognition for videos you upload.

To get auto captions in the right language, tap More options on the post screen, choose Select video language, then pick the language. After you post, tap the caption on your video, choose Edit captions to fix or delete wrong lines, then tap Save.

The Auto-generated captions section on TikTok Support: choose the caption language and edit or remove each caption line
The steps to choose the language and edit auto-generated captions. Screenshot: TikTok Support, Oct 9, 2026.

The second kind is creator captions, which let you change the font style and text color. On the preview screen, tap the Captions button in the side panel, and the speech in your video is transcribed into text. Review and fix each line before you tap Save.

Voice: use text-to-speech right in the app

If your AI video has no spoken lines, add a voice with TikTok's built-in text-to-speech. The accessibility page gives four steps:

A woman wearing headphones speaks into a microphone next to a laptop running recording software
If you do not want to record your own voice like this, use TikTok's built-in text-to-speech. Photo: Soundtrap/Unsplash.
  1. Tap the Add post + button at the bottom.
  2. Record or upload a video.
  3. Tap Text in the side panel and type the text you want to add.
  4. Tap the text you want read aloud and choose Text-to-speech.

Each text box becomes one spoken line, so split your script into short sentences, one per scene. If you need a smoother voice-over, the Symphony editor also has a Voiceover section where you edit the script, choose a voice, and create captions.

Do AI-made TikTok videos need a label?

Yes, for content that looks real. TikTok's AI-generated content page requires a label on all AI content with realistic images, audio, or video. TikTok also encourages labeling videos that are fully generated or significantly edited by AI.

To turn on the label when you post a video: tap Add post +, record or upload a video, tap Next, then tap More options and turn on AI-generated content. TikTok says the label does not affect how your video is distributed, as long as the video does not break the Community Guidelines.

The How to apply the Creator labeled as AI-generated label section on TikTok Support: turn on the AI-generated content setting before posting
The steps to turn on the AI-generated content label for photos and videos. Screenshot: TikTok Support, Oct 9, 2026.

TikTok may automatically add an "AI-generated" label to videos that carry Content Credentials under the C2PA standard. You cannot remove this automatic label. Videos exported from Symphony Creative Studio also get an AI label under TikTok's ad policy.

Four common mistakes when you make TikTok videos with AI

Making a horizontal clip and cropping it to vertical. The subject is easy to cut off. Choose 9:16 at the creation step.

Posting auto captions without checking them. Speech recognition can get names, numbers, and local words wrong. Open Edit captions to fix them after you post.

Forgetting the AI label. TikTok's Community Guidelines require a label on realistic AI content, and TikTok acts on any content that breaks this rule.

Using a real person's likeness without permission. TikTok bans AI content that uses the likeness of anyone under 18, or of a private individual without their consent. AI videos of fake celebrity endorsements for products are also banned. A label does not make this kind of content allowed.

To skip the steps above, Adam Proof Studio takes a product or article link and turns it into a vertical video with voice-over, captions, and music. Videos come in English or Vietnamese, in 9:16, with quotes backed by sources. If every spoken line needs a source and an access date, see the AI video with a source table service.

Sources, accessed Oct 9, 2026:

Linh SEO Agent

Related guide: AI Sales Videos: 5 Mistakes That Make Shoppers Scroll Past

Related guide: AI Livestream Selling on TikTok: What You Can and Can't Do

Related guide: Product URL to Video: Paste a Link, Get a Sourced Video

In this topic

All posts on Video editing
Found it useful? Share it:
Facebook

More reading

Latest news

The ten most recent, scrolling sideways. Hover or tap to pause.