If you create short-form video content in a language other than English — whether that's Latvian, Russian, Spanish, German, Arabic, or any of the other 100+ languages people actually speak — you've probably run into the same wall: TikTok and Instagram auto-captions don't work reliably for non-English speech.
They'll transcribe your Latvian TikTok as garbled English. Your Russian Reel gets skipped entirely. Your Spanish Story ends up with captions that are more embarrassing than no captions at all. Meanwhile, the English-speaking creator down the street gets auto-captions that are 95% accurate, saving them an hour of manual work per week.
This post explains how to add accurate, timestamped subtitles to your TikToks, Instagram Reels, and Stories in any language — automatically — using an SRT file workflow that takes about five minutes.
Why auto-captions fail for non-English creators
TikTok and Instagram's built-in auto-caption tools are trained heavily on English. For major languages like Spanish or French they're passable, but for smaller language communities — Latvian, Lithuanian, Estonian, Macedonian, Slovenian, Croatian, Georgian, Armenian, and hundreds of others — they either produce nonsense or simply refuse to work.
This matters beyond convenience. Research consistently shows that videos with captions get 12–40% more watch time than uncaptioned ones. Subtitles help viewers watching on mute (the majority on mobile), improve accessibility for the hearing impaired, and make your content discoverable to people who wouldn't otherwise understand your spoken language. A creator posting in Latvian who adds accurate Latvian captions — or adds an English subtitle track — immediately doubles their potential audience.
The SRT file method: what it is and why it works
An SRT file (SubRip Subtitle, .srt) is a plain-text file that pairs transcript lines with timestamps. It looks like this:
1 00:00:01,200 --> 00:00:03,800 Šodien mēs runāsim par... 2 00:00:04,100 --> 00:00:06,500 ...kā pievienot subtitrus saviem video.
Every major video editing app — CapCut, Adobe Premiere, DaVinci Resolve, Final Cut Pro, iMovie — accepts SRT files directly. You import the file, and the captions snap into place automatically, timed to the word. No manual typing, no copy-pasting, no syncing by hand.
TikTok's desktop uploader also accepts SRT files if you want to bypass their auto-caption system entirely and use your own accurate transcript.
Step-by-step: add subtitles to your TikTok or Reel in 5 minutes
Step 1 — Export or record the audio from your video
You have a few options depending on your workflow:
- Export audio from your editing app. In CapCut: share → export as audio. In Premiere: File → Export → Audio. This gives you an MP3 or M4A file.
- Use the video file directly. Raxti accepts video files too — MP4, MOV, WEBM — not just audio. Upload the whole video if it's under the file size limit.
- Record audio separately. If you script your videos, you can upload the voice-over audio directly before you shoot.
Step 2 — Upload to raxti.app
Go to raxti.app, sign in, and open the Pro Dashboard. Drag your audio or video file onto the upload zone (or tap to select it on mobile). Files up to 25 MB work on the free plan; the Pro and Business plans handle larger files by automatically splitting them into chunks and stitching the transcript back together.
Step 3 — Choose your language settings
This is the key step that makes it work for non-English content:
- Input language: select the language you spoke in the video, or leave it on Auto-detect. For less common languages, manually selecting (e.g. Latvian, Lithuanian, Arabic) gives better accuracy than auto-detect.
- Output language: this is the language your transcript, summary, and subtitles will be written in. Set it to the same language for native-language subtitles, or switch to English if you want to caption your Latvian video with English subtitles for an international audience.
Step 4 — Download the SRT file
Once processing completes — typically 30 seconds to a few minutes depending on video length — scroll to the results section and click Download SRT. You get a timestamped subtitle file ready to import.
Step 5 — Import into CapCut, Premiere, or your editor
CapCut (most TikTok creators' choice): Open your project → Text → Auto Captions → Import SRT. Your captions appear on the timeline, already synced. Edit font, size, and position to match your brand.
Adobe Premiere: File → Import, select your .srt file, drag it onto a subtitle track above your video.
DaVinci Resolve: Timeline → Import Subtitle, select your .srt.
TikTok desktop uploader: When uploading, scroll to Captions → Upload captions → select your .srt. Your manually-generated subtitles replace TikTok's auto-caption attempt.
What about Instagram Stories and Reels?
Instagram doesn't accept SRT files for upload — you have to burn the subtitles into the video itself (called "open captions" or "hardcoded subtitles"). The workflow is:
- Generate your SRT with Raxti
- Import it into CapCut or Premiere
- Style the subtitles to your aesthetic (font, color, drop shadow)
- Export the video with captions baked in
- Upload the finished video to Instagram
This is the standard workflow for any serious Reels or Stories creator — the baked-in captions look more intentional than Instagram's auto-generated ones and you control exactly how they look.
Languages that benefit most from this workflow
Any non-English creator gains something from accurate auto-captioning, but the biggest gains are for speakers of languages where platform auto-captions consistently fail:
- Baltic languages: Latvian, Lithuanian, Estonian
- Eastern European languages: Russian, Ukrainian, Polish, Czech, Slovak, Slovenian, Croatian, Serbian, Macedonian, Bulgarian
- Middle Eastern languages: Arabic, Hebrew, Farsi, Turkish
- South Asian languages: Hindi, Bengali, Tamil, Telugu, Urdu
- East Asian languages: Japanese, Korean, Mandarin, Cantonese
- Nordic languages: Finnish, Swedish, Norwegian, Danish
- Any smaller language community where platform auto-captions produce errors or nothing at all
Raxti uses OpenAI Whisper, which covers 105 languages. If you can speak it, there's a good chance we can transcribe it.
A note on accuracy
Whisper-based transcription is excellent for speech-heavy content — podcasts, vlogs, commentary, tutorials, interviews — and performs well on audio recorded in normal conditions. Background music, crowd noise, or very fast delivery will reduce accuracy. For your most important content, it's worth reading through the generated transcript before exporting, especially for names, technical terms, or numbers.
For Latvian specifically: OpenAI Whisper handles it, but dedicated tools trained on Baltic languages may produce slightly better word-error rates on challenging audio. For a subtitling workflow where you're reviewing the output anyway, the difference is usually small enough to be correctable in two minutes.
The bottom line for content creators
If you're posting TikToks, Reels, or Stories in a non-English language and not captioning your content, you're leaving reach, accessibility, and watch time on the table. The five-minute workflow above — upload audio, download SRT, import into CapCut — removes the biggest friction point for non-English creators: the absence of reliable auto-captions in your language.
You don't need to type a single word of your transcript manually.