Add Captions to Video
A free AI caption generator that works like a conversation: you describe the captions, an agent makes them.
To add captions to a video, upload it to Valmera, type "add captions", and the AI agent generates transcript-based captions for you to review against the audio. You get named presets, bundled fonts, entrance animations, per-word emphasis, and any color or position — directed through a request and checked in the preview.
Caption Your Video
Upload once, type the request, export with captions burned in. Account creation and uploads are free; editing requires a subscription.
Add Captions →How to Add Captions to a Video with Valmera
- 1Upload your videoCreate a free account at valmera.io and upload your video — up to 14GB and 3 hours. Valmera analyzes it once with visible progress and builds a word-level transcript, the foundation for word-timed captions.
- 2Type the caption requestType "add captions" — or direct the whole look in one line: "add captions in the beast preset with a yellow highlight", or "elegant captions at the top, three words at a time, in Playfair Display". The agent styles and times every caption for you.
- 3Preview, refine, exportPlay the preview and refine by replying — "make them bigger", "move them to the middle", "pop animation instead of fade". When it looks right, export a full-quality H.264 MP4 rendered from your original upload, captions burned in.
The one-time analysis of a long video takes a while and shows visible progress. After that, every caption change is just another message.
Word-Timed Captions from a Real Transcript
Valmera doesn't guess at caption timing. During the one-time analysis it builds a word-level transcript of your footage, so every caption appears exactly when the word is spoken — and captions can be grouped anywhere from 1 to 16 words per line, depending on the rhythm you want. Fast, punchy one-worders for short-form; calm full lines for a tutorial.
When the transcript gets a name or technical term wrong, correct the text in the transcript panel or request the correction in chat. Render again and check the phrase against the audio, including its timing and line breaks. Correcting caption text does not rewrite the spoken audio. The full option list is in the captions documentation.
Named presets, from Podcast to Beast
Ask for a look by name: podcast for clean talk-show lines, beast for loud high-energy captions, karaoke for word-by-word highlight captions, elegant for understated titles — plus stacked, iridescent, chrome, editorial, fashion, luxe, and impact. Every preset stays fully adjustable afterwards.
Prefer plainer captions? The classic style takes entrance animations — fade, pop, slide-up, punch, blur-in, whip, flash, rise, and drop — while the presets above animate word by word on their own. And per-word emphasis is where chat control shines: mark any word big, huge, accented, boxed, serif, chrome, glowing, or chroma-shifted, with a scale for how hard the emphasis hits.
Example caption requests: "add karaoke captions with a green highlight" · "captions in the luxe preset at the top of the frame" · "make the word FREE huge and yellow every time I say it".
Bundled Fonts, Color and Position
Bundled fonts cover the styles creators actually use: Inter Display in Black, ExtraBold, and Bold weights, Anton, Bebas Neue, Archivo Black, Poppins Black, Syne ExtraBold, Playfair Display Black, Instrument Serif, DM Serif Display, and Montserrat. Sizes run S to XL on a continuous scale, positions are bottom, middle, or top, and colors are unrestricted — including a separate per-word highlight color.
One honest limit: custom font uploads aren't supported, so check whether a bundled face meets your brand requirements before committing to this workflow. For on-screen text that isn't tied to speech — titles, lower thirds, callouts — use the separate add text to video tool, which shares the same font library.
Captions for Reels, Shorts, TikTok — and YouTube
Short-form and long-form want different captions, and both are one message away. For vertical platforms, ask for a 9:16 (or 1:1 or 4:5) reframe in the same conversation — captions rescale to the new frame — and lean on karaoke word-pop, punchy line lengths, and bold fonts. For YouTube, burned-in captions mean the text is part of the picture, with nothing for viewers to switch on. Review the uploaded result because platform resizing, cropping and compression can change its appearance.
Who reaches for this: talking-head creators captioning every upload, podcast teams styling clips, marketers whose videos autoplay muted in feeds, and educators who want lessons readable anywhere. If your priority is comprehension rather than style, the auto subtitle generator page covers the accessibility side of the same system.
Describe Your Captions
Preset, font, color, position, emphasis — one typed request. Account creation and uploads are free; editing requires a subscription.
Create an account →More Than a Caption Generator
Captions are one request inside a full agentic editor. The same upload can have its silences cut, filler words removed, music added with automatic ducking under speech, zooms punched in on emphasized words, and a color grade applied — each by describing it, and all in the same conversation as your captions. Browse every AI video editing tool to see what one chat can replace, or compare plans on the pricing page.
Frequently Asked Questions
Add Captions to Your Video
Word-timed, styled like the best creators', burned into a full-quality export. Cancel anytime.
Start Captioning →