Teach in the language your learners think in

Localizing a course is normally three painful jobs: rewrite the slides, retranslate the quizzes, and re-record the audio in a studio. Here it is a single dropdown. Set the language at creation and every generation step downstream — blueprint, slide content, section quizzes, and per-slide narration — is produced in that language from the start, so the words on screen and the voice reading them always match.

The AI is told to write as a native instructor of that language, keeping genuine proper nouns and established technical terms where locals would keep them, rather than translating everything literally. Length and quality checks are script-aware, so space-less languages like Japanese and Chinese are measured fairly instead of being flagged as “too short.”

Narration is matched to the course language automatically. The primary AI voice engine is multilingual and reads the script in its own language natively; when the backup engine is used it switches to a neural voice for that exact locale — a Hindi voice for a Hindi course, a Brazilian Portuguese voice for a Portuguese one — so the finished video never sounds like English text read with a foreign accent.

The launch set covers Spanish, Portuguese, Hindi, French, German, Arabic, Indonesian, Japanese, Tamil, Mandarin Chinese and Russian, chosen because we have strong voices in each. More languages are added as the voice quality clears the bar.