Guide to AI-Generated Subtitles for Social Video

Back in 2016, publishers told Digiday that as much as 85% of their Facebook video views happened with the sound off. TikTok is the counterexample people forget: in TikTok's research with Kantar, 88% of its users said sound is essential to the app.

Both can be true at once, because the muted viewer never went away. For that viewer, captions decide whether your first line lands or gets thumbed past, and AI-generated subtitles are what make captioning every single upload realistic.

#Subtitled video wins on the metrics that decide reach

The engagement case for captions is older than short-form video itself. A muted viewer who can read the line keeps going, and every second they stay adds to watch time. You'll see precise uplift percentages quoted for this all over the web; we couldn't trace the popular ones back to a primary study, so we don't repeat them. Add the 430 million people the WHO counts with disabling hearing loss and the accessibility argument stands on its own, before any algorithm is involved.

The mechanism is simple enough. A muted video with no text on screen has one second to communicate through picture alone, which almost nothing does. A muted video with a caption has a sentence. Platforms measure the difference as retention, and retention is what feeds distribution, so the caption track ends up being a ranking input dressed as an accessibility feature.

#Speech recognition and script-based captions fail in different places

Two mechanisms produce automatic captions, and they break differently.

Speech recognition listens to finished audio and writes down what it hears, which puts every proper noun at the mercy of the acoustics. Accuracy runs 90-95% on clear speech and drops with an accent, background music, or a product name the model has never encountered. Script-based captioning starts from the text that generated the voice, so it cannot misspell anything the script spelled correctly.

Orange takes the second route, because the script exists before the audio does. AI subtitle generation under that arrangement is a timing problem rather than a transcription problem. The distinction shows up hardest on brand names and numbers, which are the two things a viewer notices immediately when they're wrong and the two things a transcription model guesses at worst.

#Word timing comes from the voice, and the words come from the script

Orange times each caption word against the narration as it was actually spoken, so the highlight moves with the voice. The text itself never comes from a transcript. It's your script, which is why a brand name or a figure like "24,700" appears on screen exactly the way you wrote it.

#Burned-in captions are the only ones that survive a repost

Burned-in means the caption is rendered into the pixels, so it travels with the file. Download a Reel and repost it as a TikTok video and the platform caption stays behind while a burned-in one arrives intact. Sidecar SRT files are worth uploading on YouTube anyway, since the text gets indexed, but treat them as a supplement rather than the plan.

#Captions are sized for the phone, not for a fixed point size

Orange sizes captions to the frame, so a smaller export keeps the same proportions instead of shipping text that runs off both edges.

The one-word style shows a single word at a time and can't wrap, so it scales down until the longest word in the narration fits. Position is bottom-center, held clear of the strip where the TikTok caption box and the Reels action rail sit. Captions placed by eye in a desktop editor land underneath that furniture surprisingly often.

#Non-Latin scripts need the font embedded or the export ships boxes

A subtitle renderer with no glyph for a character draws a tofu box, and it does this silently. Orange carries the right fonts for CJK, Arabic, and Devanagari scripts through the render itself, so a Japanese or Hindi video ships its own typography rather than depending on whatever the encoder happens to have.

Numbers matter for anyone posting in 34 languages. In a Czech or Slovak video the voice says a figure the way a native speaker would, while the caption keeps the written number rather than a wall of spelled-out words.

#The highlight should stop with the voice, not with the scene

Word-by-word highlighting has a failure mode nobody notices until they look for it: the last word of a line keeps lighting up while the visual holds for another second. In Orange the highlight finishes when the voice does, and the completed line stays on screen until the cut, so it's still readable through the outro.

Five styles ship today: bold, minimal, animated, one-word, and boxed. For short-form, the animated word-by-word fill earns its keep, because the moving highlight gives a muted viewer something to track. Longer educational cuts read better with full sentences. Auto-subtitles run on every render regardless of style, with no extra step to trigger.

#Best practices worth keeping

1. Put captions on everything. Every social video, no exceptions.

2. Use high-contrast text. White with a dark outline or a background bar survives any visual underneath. Thin light-colored type does not.

3. Cap it at two lines. More than that and a phone viewer reads instead of watching.

4. Time it tightly. Words should appear as they're spoken. Even 200ms of lag reads as broken.

5. Check the bottom edge. Platform UI eats the lowest strip of the frame, and the amount varies by platform and aspect ratio.