I answer members so mods don't have to.
Your team writes down the answer to a repeat question once. Next time anyone asks, I reply using your wording and cite the entry. If I don't know the answer I say so — I never make things up.
Orbital
Every answer is grounded in your team's own wording, cited back to the entry it came from. Works in twenty languages out of the box. The free tier covers a hundred AI calls a month with no feature locks.
A typical exchange.
/verify start and answer the question I ask — it's always fresh, drawn from the server rules. Once you get it right the role is granted straight away. [verify]The [verify] tag is a citation — it's the KB entry your mods wrote. I never invent claims that aren't there.
All five are on every tier, including free.
Your team writes down the answer to a repeat question once. Next time anyone asks, I reply using your wording and cite the entry. If I don't know the answer I say so — I never make things up.
The moment a member opens a ticket, I read the subject and try to answer using the same knowledge base. Staff only see the ones I couldn't close. It roughly halves triage time.
KB entries are tagged per language. Set the server default, let individuals override. I ship with 20+ locales from day one — French, Spanish, Japanese, Arabic, the full list.
Basic automod, raid-spike alerts, and a weekly summary of every moderation action — posted to a channel of your choice, written as a readable paragraph, not a spreadsheet.
Every tier shows you exactly how many AI calls you've used. Denials link you to the reset time. No hidden caps, no surprise bills, no feature-locked parts of the product.
Every tier ships the same product. Paying gets you more AI calls each month and a few background features (longer answers on Pro, BYO-Gemini-key) that would cost too much to run at the free tier's scale. The commands, the dashboard, the archetypes, the integrations: all the same.
Running more than one server? There are discount packs: three servers for $24 a month, ten for $70, twenty-five for $150. Any tier can be cancelled from the Stripe portal at any time.
See the full privacy page and terms.
I never train any models on your data. I never sell anything to anyone. The only third parties I talk to are Google, when someone asks me a question that needs the AI, and Stripe, when someone upgrades. Both see only the minimum they need to do the job.
If you decide Orbital is not for you, remove me from your server and everything tagged with your server's id is wiped on the next cleanup pass. Email support and it happens immediately.
No. You click Add, pick your server, and then answer three quick questions on a setup page — an audit channel, a staff role, a welcome message. That's it. I run on our servers, so there is nothing for you to install or configure.
Every answer I give is tied to an entry your moderators wrote, with a visible citation. You can click through and check the source any time. If the knowledge base does not cover a question, I say so instead of guessing. Your moderators should still read my replies before acting on anything serious, especially moderation decisions.
There isn't one. A hundred AI calls each month is enough for a small or medium server asking a few questions a day. Busier communities tend to hit the limit, and three dollars a month lifts that to two thousand calls. Nothing in the product is hidden behind a paywall.
Nine times out of ten it's role order. In Discord, a bot can only touch roles that sit below its own in Server Settings → Roles, so drag my role above any role I need to assign (verified members, ticket openers, role-menu picks). The other common one is that the channel I'm meant to post in doesn't grant me View Channel + Send Messages. If none of that helps, open the dashboard — I'll tell you exactly which permission I'm missing.
Yes, and plenty of servers do. I don't touch moderation commands those bots already own, so there's no clash — you get knowledge-base answers, tickets, and welcomes from me, and everything else from whoever you already trust. If two bots end up replying to the same thing, tell me to stop answering in that channel from the dashboard.
As little as possible. I store your knowledge base entries, your ticket history, your per-server settings, and a log of moderation actions. Regular member messages never touch our servers unless automod removed one. When someone asks me a question, Google Gemini sees that question and the matching knowledge base entries, and nothing else. I never train any models on your data, and I never sell anything to anyone.
Kicking me stops new writes immediately. To delete the existing data — knowledge base, tickets, audit events, settings — email support@orbital.app from the Discord email of the server owner and we'll run the purge and confirm. We're building an automatic deletion pass; until it ships, the email route is how we keep the promise honest.