Auto-generated subtitles

Creatomate transcribes the speech in a video or audio element and displays it as synchronized captions in the rendered video. There is nothing to set up and no external service needed; transcription is built into the platform.

Adding subtitles

Select the video or audio element you want transcribed, and click Add Subtitles in the toolbar above the canvas. This creates a subtitle element that is already linked to it and styled to read over the video.

A subtitle element is a text element whose words come from the transcript instead of being typed. Select one and the properties panel titles its section Subtitles, where Source names the video or audio element being transcribed. Make sure that element plays at the same time as the subtitle element; the subtitles cover the part of the speech during which the subtitle element is shown.

Everything below it is the usual text properties: font, size, spacing, alignment, and the fill, stroke and background that decide how the text is drawn. A subtitle element is styled exactly like any other text element.

Choosing an effect

How the words behave is chosen in the Animation tab, under the Caption filter. Every effect is a live preview tile.

Below the tiles, a strip carries the four settings that shape the chosen effect, so you can tune them against the one you are watching:

• Split: Whether the effect applies to each line, word, or letter, based on its timing in the speech. Clicking the selected one again turns it off, which applies the effect to the whole sentence at once.
• Max. length: The maximum number of characters shown at once. Use a low number to display just a few words at a time, or auto to split sentences automatically. Setting it to 1 gives the one-word-at-a-time style common on YouTube Shorts, TikTok and Instagram Reels.
• Placement: Show the whole sentence and accent each word as it is spoken (Static), reveal each word in place while keeping the whole sentence's layout (Animate), or show only the words spoken so far, re-flowed to the text alignment (Align).
• Color: The accent color the effect uses, such as the color of the currently spoken word.

Every change plays back on the template, so you hear and see the result as you make it.

Any of these can be made dynamic from the Dynamic row at the top of the properties panel, so your automation can change them per render.

What does it cost?

Transcription is free for videos rendered at 720p or higher. For lower resolutions, a small transcription fee applies only where it exceeds the render's regular credit cost. See How are credits calculated? for details.

Related workflows

Subtitles work especially well with AI voice-overs: set the voice-over element as the transcription source, and the generated speech arrives with exact word timings. And if you already have subtitles, or want full control over the wording, you can provide your own transcript through the API instead.