Video editing guide

Add Titles and Text Overlays to a Video Online

A title should be easy to read, spelled correctly and on screen for the right moment. Add a timed text layer yourself or ask the assistant with the exact words and placement, then check it while the video plays.

Decide what the text is for

An opening title, a short instruction and a label pointing at something all need different timing. Keep the words short enough to read while the video carries on; a full sentence over a fast demo makes people choose between reading and watching.

Choose the position before adding it. The bottom of the frame may already hold burned-in subtitles, a product label or a player’s controls, so check the whole time the text will be on screen.

Add a text layer with a start and an end

Move to the right point on the timeline and add text, or describe it in a request. Type the exact words, drag the text where you want it in the preview and resize it with the corner handle. Set its start and end within the current project length.

Use project time, not the time in your original file. If you cut the opening, the beginning of the edited video is a different moment in the source.

Add “Before” at the top from 00:02 to 00:05 and “After” at the top from 00:08 to 00:11, white on a dark background.

Make it readable on the final frame

White text on a dark background works over most footage. If you pick another colour, test it over both the brightest and darkest moments while it is on screen.

Look at it at the size people will watch. Text that seems fine on a laptop can be too small on a phone. Split a long line into two timed layers rather than filling the frame.

Check when it appears and disappears

Play from just before the start to just after the end. The words should arrive when they become relevant and stay long enough to read. If you cannot see the text in a paused preview, check the playhead is inside its time range.

After cutting or reordering, re-check every title. Edit the layer itself rather than adding a second copy, so the same words do not appear twice.

Titles are not automatic captions

A text layer can hold a title, a label or a line of speech you type in. The beta does not transcribe audio or build a full set of captions for you.

If you type spoken lines yourself, check them against the audio and give each line its own readable time. The assistant does not receive a transcript of what is said.

Export with the text burned into the picture

Text becomes part of the exported frames, so viewers see it without a separate subtitle file. After exporting, check the first, middle and last frame of each title and read every word.

Keep the original video and a project backup if the wording might change later. The export contains the rendered text; the project keeps it editable.

Frequently asked questions

Can I show text for only the first few seconds?

Yes. Set the layer to run from 00:00 to the time you want, or include those times in your request.

Will the text be in the exported video?

Yes. It is drawn into the frames when you export. Check the downloaded file for spelling and timing.

Can it write subtitles from the audio?

No. Automatic transcription and captions are not in this beta. You can type timed text layers yourself.