SeekDeep — Text-to-Speech

Text-to-Speech

▸ Piper · local · offline

HOW IT WORKS

If prompted above, click Install speech engine (one-time, ~30 MB).
Pick a voice on the right and hit download.
SeekDeep fetches the Piper model (~60–110 MB) and switches to it automatically — no restart.
Type text in the middle and press Speak to hear it right here.

Downloading a voice also writes SEEKDEEP_TTS_PIPER_VOICE + SEEKDEEP_FEATURE_TTS_VOICE=on to your .env, so it survives a restart.

IN DISCORD

Join a voice channel, then mention the bot:

@SeekDeep tts join
@SeekDeep tts say <text>
@SeekDeep tts leave

⚠ The Speak test above works immediately, but the Discord bot reads the TTS flag and the voice-channel intent only at startup — restart the bot once after enabling a voice for the tts commands to work in Discord.

Discord voice playback uses a bundled FFmpeg (ffmpeg-static) — no install needed. If tts say ever reports an FFmpeg error, install FFmpeg on the host PATH.

Test your voice

▸ synthesized locally · plays in your browser

SYNTHESIZED AUDIO

SeekDeep ▸ Download a voice on the right, then type below and hit Speak — the audio shows up here.

engine Piper tip Ctrl+Enter to speak