Assamese AI voice generator and voice cloning

• Reference clips for Assamese are capped at 12 seconds.
• Assamese has letters Bengali does not — ৰ (ro) and ৱ (wo) — and they must be typed as the Assamese characters rather than their Bengali lookalikes. Substituting the Bengali forms changes the sound.
• Roman-script Indian-language text and English words mixed into an Indian-language script are normalised before generation. Without that step, models trained only on native-script audio tend to skip Latin-script words silently rather than pronounce them.

These are the exact prompts Mayantar shows you in the recorder. Each one is timed to land inside the 12-second limit at a natural pace, and each covers something cloning models need: relaxed conversational prosody, spoken digits and dates, and a question with rising intonation. Read one aloud in a quiet room and you have a usable reference clip.

1. নমস্কাৰ, এইটো মোৰ কণ্ঠস্বৰ আৰু আজি মই এটা সৰু নমুনা ৰেকৰ্ড কৰি আছোঁ। মোক কিতাপ পঢ়া ভাল লাগে।
2. আহক কেইটামান সংখ্যা কওঁ: 1, 2, 3, 4 আৰু 5। মোৰ সভা ৰাতিপুৱা 9:30 বজাত, 15 মাৰ্চত আছে।
3. আপুনি কেতিয়াবা ভাবিছেনে যে এটা সাধাৰণ ৰেকৰ্ডিঙে কেনেকৈ কাৰোবাৰ কণ্ঠস্বৰ ধৰি ৰাখে?

More detail on what makes a clip pass or fail is in our guide to recording a reference clip.