Korean ad with an English brand name: set Sume TTS language to ko
Send a Korean script with language set to ko; the default is English. A brand name in Latin letters is read by the chosen language, so listen and fix it.

For a Korean ad that includes an English brand name, send the whole script in one request with language set to the Korean code. Sume TTS defaults to English when the field is missing, which makes the Korean text read badly. Then listen to the brand name; the engine reads Latin letters under the chosen language, so a pronunciation fix may be needed. I did not verify the exact language code list, so check the OpenAPI schema for the accepted value before you ship.
Three settings that matter
The TTS reference lists language as required for non-English requests and shows English as the default. It also defines confirm_language_mismatch for the case where the voice and language differ, and pronunciation_dict_id for terms that need a particular reading. Together these are the whole toolkit for a mixed-language script.
| Field | What it does | Use in a Korean ad |
|---|---|---|
| language | Sets the language of the script; English by default | Set to the Korean code from the OpenAPI schema |
| confirm_language_mismatch | Accepts a voice whose language differs from the script | Use only when you meant it |
| pronunciation_dict_id | Applies a dictionary of preferred readings | Add the brand name with its intended sound |
Captions for the same ad
Burned-in captions for Korean follow their own rules in the Video captions docs, which describe Hangul handling for caption styles. If the brand name must appear in Latin letters in the caption while the voice says it in Korean, send the exact words in script_text so the caption shows your copy; the speech timings still decide when each cue appears. Check the video captions docs for the rules before relying on a specific style.
A short test before the batch
Produce one 100 character sample that contains the brand name. At $0.0475 per 1,000 characters that is $0.00475. Play it, and if the name sounds wrong, add the term to a dictionary and run it again for another $0.00475. Only then send the full script, whose cost is the same rate times its length. A 400 character ad is $0.019.
What can go wrong
If you leave language out, the request reads in English and Korean text comes out wrong. If you choose a voice made for another language, you get the mismatch response; confirming it is allowed but not advised for an ad. If the brand name is read in Korean sounds when you wanted English ones, spell it as it should sound in the dictionary entry. Keep the script free of stray symbols and emoji, because they can be read aloud or skipped, depending on the engine.
Sources
Related posts
More in Developers
- Korean voiceover: 1,200 Hangul characters cost 5.7 cents, so use NFC
Sume TTS bills $0.0475 per 1,000 transcript characters, so 1,200 Hangul characters cost 5.7 cents. Decomposed text can count two or three times more.
- Lint a video against TikTok, Reels, Feed and Shorts limits in Python
A 24-line Python check for duration, file size and average bitrate against the limits TikTok, Meta and YouTube publish, plus where Sume jobs fit before it.
- Lint a Video Router request offline: reference limits in Python
Check reference_image_urls (10), reference_video_urls (3), reference_audio_urls (1-5) and the video_url rule in Python before POST /v1/video-router/generate.
- List Sume's video endpoints from the OpenAPI JSON in Python
Pull the OpenAPI snapshot, print every /v1/videos and Video Router operation, and know which copy is the source of truth. A short Python script.
Written by Sume