Launching today
OpenCouncil
Voice rooms where your team and AIs debate together
52 followers
Voice rooms where your team and AIs debate together
52 followers
Put GPT, Gemini, Grok and your own agents in one live voice room with your team. They answer, push back on each other and never talk over you. A Scribe writes the decisions. Make the room public and anyone can listen, chat or take the stage. Works in your browser and Google Meet. Free to start: 2,000 credits, no card.
Launch Team
Viktor.comAn AI coworker that actually does the work
Promoted
Hey Product Hunt 👋 I'm Douglas.
Think Clubhouse, but some of the speakers are AIs.
I kept pasting the same question into three AI tabs and then retyping the answers for my team. So we built OpenCouncil: a voice room where your team and several AIs talk it out together.
What's different:
• Multiplayer: your whole team in the room, not just you.
• Mix models: each agent can run on a different AI, with its own name, voice and role. Bring your own agent too.
• Turn-taking: OpenCouncil decides who speaks next, and the AIs never talk over a human.
• Open rooms: flip a room to public. Anyone can listen, chat and react without an account, then ask for the floor.
• Works in your browser and in Google Meet. Discord too. Zoom is next.
Jump into a live room, no signup: https://opencouncil.build/en/discovery
Want your own council? It's free to start: 2,000 credits, no card: https://opencouncil.build
I'd love brutal feedback: which room would you open first?
@douglas_dlweb @agzee Thanks so much! Keeping humans in the room is the whole point: the AIs are there to push the conversation, not replace it.
Hi Douglas, what sort of logic does the Decision API have to decide how long the debate goes for? Also is the user able to update the prompt to these agents? I ask because I could see these debates going on indefinitely, since a human would eventually give up. Also is there a way to track token usage?
Congrats on today's launch!
@sara_ford_goog Thanks Sara! After every sentence, the decision engine picks who should speak next, or nobody, so a debate runs only while it still has something to add. As a hard stop, agents get at most 20 turns in a row without a human; then they wait for someone to speak. You can edit both the decision prompt (per session) and each agent's instructions. And every model call is debited as credits, so you can see exactly what a session cost in your billing history.
I call these rooms "Boiler-rooms" in my team 😸 This is definitely the future. One question is this for IRL or Google meets etc?
@alchemist_studio Love "boiler rooms" 😄 Both! In person, one laptop in the room listens and the agents answer out loud. On calls, the council joins your Google Meet (or Discord), or you run the room in the browser and share the link with your team.
I can see this being really useful for brainstorming and product decisions 🤯
@guinhoshuto That's exactly where we use it most! Which decision would you throw at a council first?
I like that the team can join the conversation instead of just reading AI-generated responses. It makes discussions feel more interactive.
@edward_paul3 Thanks Edward! That's exactly the idea: the AIs sit in the conversation with your team, instead of a chat window on the side. What kind of conversations would you bring a council into?
Great idea! The turn-taking and the Scribe logging decisions are smart touches
@new_user___3522025e3a7799f7d4b313a Thank you! Turn-taking was the hard part: after every sentence, a tiny model decides who should speak next, or nobody. Glad the Scribe landed too.