Discord accessibility, in both directions
Give people more ways to take part in a Discord call. Mimiq speaks typed messages and turns speech into captions, so members can choose how they join the conversation.
Add Mimiq to your serverFor members who type instead of talk
Auto-read: type in VC chat, be heard
With auto-read on, eligible messages in the connected voice channel’s own text chat are spoken aloud to everyone in the call, no commands, no interruptions to the flow of conversation. Whether someone is mute, has a speech impairment, has no working mic, or just can't talk right now, they're part of the conversation in real time.
A voice of their own
Members aren't stuck with one generic voice. Everyone picks a personal default with /setvoice, and with /clone, a member can even use an AI clone of their
own voice, made from a short recording, with consent confirmation built in.
For members who read instead of listen
Live captions for voice channels
A member with Manage Channels can run /captions start and Mimiq transcribes the voice
channel into its text chat as people speak, labeled by speaker. Deaf and
hard-of-hearing members follow the call as it happens, and everyone else gets a
scrollback of what was said.
Run both: the two-way bridge
Turn on auto-read and captions in the same voice channel and it becomes fully two-way: typed messages are spoken into the call, and spoken words are captioned into the chat. A member who neither speaks nor hears participates in the same channel as everyone else. Nobody gets sent to a separate "accessible" corner.
Questions
Is this free? Accessibility tools usually are not.
Free includes unlimited monthly TTS and typed-message translation, 3 custom voices and 60 caption minutes each month. Automatic caption start and live caption translation require Pro.
Can someone who cannot speak have their own voice?
Yes, and this is the part people love: with /clone they can create a custom voice from a recording they own or have permission to use and set it as their default. They can use it with /tts or auto-read, subject to the server’s settings.
How accurate are the live captions?
Captions are powered by Deepgram, one of the leading speech-recognition engines. Clear speech in a reasonable-quality call captions very well; heavy crosstalk or poor mics lower accuracy, as with any transcription.
Does the person using TTS or captions need special permissions?
TTS and auto-read follow the server’s channel and role rules. Starting or stopping captions requires Manage Channels. Changing server settings in the dashboard requires Manage Server.
What does setup look like?
Add the bot, and it works out of the box. Auto-read is one toggle in the web dashboard. Captions start with one command. There is no configuration wall to climb.