Resonant + Discord
The helpful reply
you didn’t have time to type.
Discord runs on long-form replies — explanations, feedback, writeups, the thoughtful paragraphs that make a community worth showing up to. Those paragraphs are the ones most likely to get skipped, because typing them is work.
Resonant lets you say the full answer and drop it into any channel. On-device speech recognition. No cloud audio. The same Mac app works in the Discord desktop client, Discord in the browser, and every other text field on your machine.
Used by community members at
Typed vs. dictated
Same channel. Same moment.
A reply the community actually uses.
Reply in a help channel
“check your config file”
“You’re probably hitting the same thing I ran into last month — if you installed the CLI via Homebrew but then ran the init script as sudo, the config ends up in /var/root instead of your home directory, and the next invocation can’t find it. Check whether ~/.config/thetool exists. If not, delete the copy under /var/root, re-run init without sudo, and it should work. If the file is there but empty, that’s a different bug and I can walk you through it.”
Community update post
“update: v2 out”
“Quick update on where we are: v2 is live as of this morning. The big changes are the new plugin API, the rewritten settings panel, and the fix for the long-standing token refresh bug. I want to flag that the plugin API is technically in beta — the shape won’t change but the docs are still catching up, so expect rough edges. If you hit anything weird, post in #bugs with a repro and I’ll triage today and tomorrow.”
Feedback on a project
“cool, couple of notes”
“Really liked the direction — especially the way the onboarding flow puts the demo before the signup. That ordering is rare and I think it’s going to convert. Two notes: the pricing page feels heavy because each tier has five bullet points and they’re all similarly weighted, so the eye doesn’t know where to land. I’d star the two or three differentiators per tier and demote the rest. Also, the top-nav search feels like it wants to be a command palette — if you go that direction, it’s a huge upgrade.”
The insight
Communities
select for short replies.
Voice reverses that.
Every async community runs on an invisible cost function. Writing a reply has a fixed overhead — thinking, typing, editing, proofreading. The shorter the reply, the lower the overhead, so the reply you actually send is almost always shorter than the reply you wanted to send. Over hundreds of messages, the community ends up selecting for terseness, even when terseness is the opposite of what helps.
Discord is where this shows up most. It’s the place where long-form explanations, project feedback, and help-channel answers live. The communities with the best writeup culture are also the communities that burn out their most active contributors — because writing four thoughtful replies a day at 40 wpm is a part-time job nobody signed up for.
Voice dictation flips the incentives. The full explanation now costs less than the one-liner. The feedback paragraph takes thirty seconds instead of five minutes. The writeup that would have lived only in your head gets posted, and the next person who searches the channel finds it.
You know the answer. You also know the answer takes four paragraphs to explain properly. You send “check your config” instead and hope they figure out the rest. The knowledge stays stuck in the people who already have it.
Helpfulness shouldn’t be bounded by wpm.
When replies cost less to write, more of them get written, and the average reply gets longer and better. Help channels start answering the real question. Feedback stops being thumbs-ups. New members get welcomed with paragraphs instead of emoji.
The community compounds.
Where it fits
Six places in Discord
where voice changes what gets said.
Help channel answers
The most useful answer in a help channel is usually the one with enough context to prevent the next three follow-ups. Voice makes that answer cheap. You talk through the thing you already know, and the next person with the same bug finds a real explanation instead of a one-liner.
Long-form feedback
Design crit, code review, startup advice — the stuff that actually helps someone is a few paragraphs, not a thumbs-up. Dictating feedback lets you give the full thought without it turning into a chore that you defer and then forget.
Organizing messages
Community mods and team leads spend huge chunks of their week writing announcements, thread pins, and rules updates. Voice turns the twenty-minute message into a two-minute one, which means updates actually ship instead of sitting in drafts.
Sharing what you learned
The best communities are the ones where members write up what they figured out. Those writeups get skipped because typing them is work. Say the story out loud into a thread — the next person searching for the same problem will thank you.
Voice channel follow-ups
You had a voice call, something important came out of it, and now someone needs the summary. Dictate it into the channel while the call is still fresh. The five people who missed the call get the context without you rewriting the conversation from memory.
Welcoming new members
Good onboarding in a Discord means an actual, human welcome with the things a newcomer needs to know. Voice makes personalized welcomes sustainable instead of replacing them with a pinned auto-message nobody reads.
Architecture
Your DMs stay your DMs.
Your audio stays on your Mac.
Resonant does speech recognition locally on Apple Silicon. Audio is captured, processed in memory, and turned into text without ever touching a server. The only thing that leaves your Mac is the final text — the same bytes that would have left if you typed the message yourself.
That’s important for Discord specifically. DMs contain personal stuff. Private servers contain work-in-progress projects, early access betas, and conversations between people who trust the server to be small. A cloud dictation tool would route every message through a third party’s transcription pipeline. Resonant doesn’t.
There’s no Discord bot, no bot token, no permissions to configure in server settings. Resonant is a macOS utility that writes to whatever text field has focus, which means it works in every Discord server you’re in, without any server owner needing to install anything.
Free. Local. Works in every Discord server.
Be the long-reply person
without the long-reply effort.
Real answers in help channels, real feedback on projects, real writeups in threads — at speaking speed.
Requires macOS 14+ · Apple Silicon
Related