Lerche TTS

TTS people actually listen to

• Lerche gives each member their own personalised, natural-sounding voice through ElevenLabs and Fish Audio, making typed messages feel like a real part of the conversation.
• When someone cannot speak, does not feel comfortable talking, or simply prefers to type, their messages can easily get missed. Lerche is built to help people pay attention, respond, and include them in the call.
• Members can choose their own model and voice for each server with /personal-set-my-voice, alongside a default voice chosen by the admins. Optional spoken names help everyone follow who is talking.

• I needed a TTS bot like this at one point. Something people would actually listen to, that made typing in a voice call feel comfortable and helped me feel included.
• That is what I am trying to make with Lerche.
• She helps people be heard.
• She's quick.
• Most importantly, she is built by a real Discord user who needed her too.

If you install Lerche, thank you. If she helps you, I'm genuinely glad!

• An admin can run /tts-install for guided setup.
• Then join a voice channel and send a message in the configured TTS channel.
• Voices and other settings can be adjusted afterwards.
• Use /help if you need a hand.

Prefer your usual text channels? Admins can enable **`/t` room mode**. Type `/t` followed by your message, and Lerche reads it in your current voice channel.

Messages queue smoothly, mentions and links are cleaned up before being spoken, and she leaves after a configurable period of inactivity.

• Voice generation costs money, so each server has a TTS allowance.
• Lerche warns you as it gets low and pauses when it runs out.
• Admins can check the balance with /info-token-count.
• If you need more, supporting Lerche on Patreon helps cover usage and development.

• Role controls — allow everyone or choose who can use TTS.
• Optional jingles and confirmation replies — adjust how much Lerche announces herself.
• Playback controls — stop playback, clear the queue or move her to another voice channel.