Thai audio and video to text

Central Thai and regional speech

Central Thai, the Bangkok standard used in news and education, is the easiest input. Northern (Kham Mueang), Isan (Northeastern) and Southern Thai are related but differ in tones and vocabulary, and are harder for the model. Speech with a lot of English words, common in tech and business, is also weaker. Formal and casual particles such as ครับ and ค่ะ are written as spoken.

Thai script and the lack of spaces

Thai is written without spaces between words. Spaces mark pauses or the ends of sentences, so a transcript can look like long unbroken lines. Tone marks and vowels sit above and below the consonants, and a missing or wrong one changes the word. Editors and players may break lines in odd places, so check the subtitle wrapping.

What speakers of Thai use it for

Recorded university lectures, meetings and calls for work, interviews and podcasts, and captions for YouTube, TikTok and livestream clips. Journalists and researchers use transcripts to quote accurately. Since there are no spaces, a text version also makes it easier to search for a name than scrubbing through audio.

Which model works best for Thai

Thai is a lower-resource language for Whisper-family models, and the small model is noticeably less reliable here than for the major languages. Use clear, single-speaker audio, keep clips short, and plan to edit a good deal. For the best result, self-hosting with large-v3 helps.

On tryscriptorium.com the no-account preview and regular accounts run the small model (the larger medium and large-v3 models run when you self-host Scriptorium). In an account, pick Small in the model menu rather than Auto: Auto prefers the fastest, least accurate model for long files or a busy queue.

Thai is one of the weaker languages for the small model. Expect more wrong words than in Mandarin or Spanish, especially with noise, music or several speakers. Use short, clean clips and proofread everything you publish. For anything long or important, self-host with large-v3.

Subtitles and captions in Thai

Thai subtitles should be broken at phrase boundaries, since there are no spaces to rely on, and kept to about 35 to 42 characters per line. Some players wrap Thai in the middle of a word, so test the file in your player. SRT and VTT are UTF-8, so Thai script and tone marks display correctly in VLC and YouTube.

The uploader is set to Thai

The upload box on this page transcribes as Thai, so the recogniser does not have to guess the language from the first seconds. If the recording is in another language or switches languages, use Audio to text, which detects the language automatically. Automatic transcription is a draft and no accuracy figure is promised: check names, numbers and anything you will rely on.