Voice & Audio

TTS, STT, and music generation with a real free allowance.

Monthly character cap suitable for short voiceovers, not a catalog.

High-quality TTS. Free tier is enough to voice a demo, not a channel.

Daily credits for full songs with vocals. Enough to demo, not to ship a catalog.

Text-to-song. Fastest way to a verse-chorus demo with a vocal. Use it for scratch tracks, not masters.

Monthly generation cap. Stronger mix quality than most free song tools.

Second music generator worth keeping. Use when Suno’s vocal is wrong and you still need a free take.

MIT. Unlimited once the ggml model is on disk.

Local speech-to-text. The transcription plan that still works when Gemini is blocked and you cannot upload audio to a US API.

MIT. On-device TTS. Voices are separate downloads.

Fast local text-to-speech. Good English/other voices. Treat Burmese as unsolved — do not ship Piper as native MM narration.

Apache-2.0 82M model. Runs on CPU. Unlimited once downloaded.

Small local TTS that punches above its size on English. The no-GPU voiceover option next to Piper.