.png){kind=logo}
They ask in the group.
The agent replies in the thread.
Drop in the bot token once and the agent handles your midnight waves. It cites every answer, escalates delicate threads, and keeps moderators answering only what truly needs a human.
Type a slash.
Tap a command to see the agent respond with a grounded answer and @mention etiquette. Your moderators review flagged questions and downvotes in the dashboard instead of catching up on 600 messages.
One bot. Twelve moves.
Every other Telegram AI stack forces you to pick one bot and glue three more on top. Combot can't answer. Chatbase can't moderate. Livegram can't be proactive. Guardian can't be brainy. We do all twelve — one dashboard, one bot token, one knowledge base.
Six moves. How the agent answers.
Each capability is configurable per group or DM. The agent stays inside your knowledge and tags you when it shouldn't answer.
Threaded replies in busy groups quote the question, @mention the asker, and keep the chat skimmable during launch spikes.
In groups the agent waits for @mentions or matched FAQs; in DMs it replies instantly, so no one feels ignored.
Pulls each member’s first name from Telegram, keeps tone on-brand, and never slips into generic bot phrasing.
Map launch FAQs, policy reminders, or ticket links to slash commands in seconds. Mods answer once; everyone reuses it.
Listens for FAQ patterns and steps in without the @mention, then stays silent if it can’t back the answer with a source.
Detects the language of every incoming message and replies in that language automatically. Your knowledge base stays in one language — the agent handles the translation on the fly. Tested across English, Spanish, French, Yoruba, and 50+ more.
Six more. How the agent runs the community.
Not just answers — the full stack of moderation, safety, and workflow that keeps busy groups healthy without another bot.
Configurable warn ladder (defaults 3 → mute, 5 → kick, 7 → ban) with a keyword filter that auto-deletes banned words and warns the sender. Every action lands in the audit log.
Every new member is checked against the free public CAS spammer database, optionally challenged with a tap-the-button captcha, and auto-kicked if flagged as a Telegram bot. Blocks known raiders before they enter.
When the agent hits the edge of its knowledge, users tap a "Talk to a human" button. Admins get a DM ticket and reply with Telegram's native Reply gesture — the bot relays the answer back into the chat, no dashboard round-trip.
Author rich cards with title, body, image, and grids of inline buttons in the dashboard. Trigger on slash commands, on member join, or from another menu's sub-menu button. Users navigate in-place, no chat spam.
Schedule text or GIF messages to any group the bot is a member of, every 2 / 3 / 4 / 6 / 12 / 24 hours or weekly. Perfect for announcements, product tips, or launch pings.
Users tap under each AI response. Votes flow to your dashboard — satisfaction %, downvote list to review, direct signal into what to add to your knowledge base next.
Three steps. Five minutes.
Create a bot with @BotFather, paste the token in the dashboard. We verify scope, TLS, and webhook health before it goes live.
Same knowledge that powers your website. Set tone, sign-off, escalation rules, and which slash commands the agent exposes.
Add the bot to your group, pin the welcome message, and watch the same five FAQs stop landing in your mods’ DMs.
Built for the people
already shipping.
Communities, courses, drops, betas, internal teams, agencies. Six rooms where the same five FAQs vanish and moderators get their time back.
Token-holder groups asking “is this real?” all day. The agent answers from your authentication and roadmap docs, so holders stay calm and holders stay in.
Students drop the same “what’s due?” question 200 times. The agent points them to the syllabus while mentors focus on coaching.
Launch waves cry “when is the drop?” and “am I on the list?” The agent handles both at scale with timestamped answers.
Beta testers report the same bug five times. The agent flags known issues, links to the GitHub thread, and only pings engineers for new ones.
HR, ops, and finance Q&A handled from SOPs so no one pings leadership at 11 PM asking where to file expenses.
White-label the agent inside client groups. Status updates, retainer questions, and handoffs read like they came from the account lead.
Your data, your rules.
Token-isolated. Grounded in your own knowledge base. Scoped to Telegram API permissions you approve. GDPR-aligned handling, audit trails on request.
- ◉ The same five FAQs stop landing in moderator DMs every week.
- ◉ New members get welcomed, FAQ'd, and oriented while you sleep.
- ◉ Group chat stays clean. The agent only replies when it has something useful.
- ◉ Every AI reply carries a 👍 / 👎 — you see satisfaction, downvotes, and unanswered questions in the dashboard.
- ◉ Bot token stays encrypted at rest and tenant-isolated. Rotate it anytime from the dashboard.
- ◉ The agent only answers from what you uploaded. Never the public internet.
- ◉ Webhooks monitor themselves. If Telegram drops, we retry and alert you—no midnight resets.
- ◉ Pause, mute, or remove the agent in any group from the dashboard, instantly.
New member waves.
Same five FAQs.
The agent picks them up so moderators can stay human. Quiet when it doesn’t know, loud (and cited) when it does.
Stop scrolling the same Q&A.
Pin the agent.
Paste the token, train the voice, pin the welcome. The agent answers every group while you focus on what only you can do. Need procurement or security reviews? Our team runs onboarding alongside your admins.
plans from $20/mo · see all plans →