One slogan in English, Korean and Japanese: three TTS jobs, cost

Voicing one 120-character slogan in English, Korean and Japanese takes three Sume TTS jobs at $0.00570 each, $0.01710 total at $0.0475 per 1,000 characters. Set language on every job (en, ko, ja) and send each language's own text.

Cost by language count

Assuming each version is about 120 characters. Real counts differ by language, so recompute with the final strings.

Loop over languages

Use one key per language so a retry reuses its own job.

versions = {
    "en": ("@narrator-en", "Light on the road, strong in the rain."),
    "ko": ("@narrator-ko", "가볍게 걷고, 비에도 든든하게."),
    "ja": ("@narrator-ja", "軽やかに歩き、雨にも強く。"),
}
for lang, (handle, text) in versions.items():
    body = {"transcript": text, "avatar_handle": handle, "language": lang}
    r = requests.post(f"{API}/v1/tts-1.0/generate", json=body,
                      headers={**H, "Idempotency-Key": f"slogan-{lang}-v1"})
    r.raise_for_status()
    print(lang, r.json()["data"]["job"]["id"])

Gotchas

The handles above are placeholders; list your real avatars and confirm each has a ready voice for its language. Omitting language defaults to English at the provider, which would read Korean text badly.