Each feature does one job well. They all run through a shared account fleet with common proxies and rate-limits.
Get your message in front of the right people, at scale.
Find audiences. Contact them. Convert them.
Mass-invite a username list to your channel or group
Distributes a list of target usernames across your account fleet, and each account adds its share to the destination channel/group. Handles per-account daily caps, contact-add-before-invite (raises success rate), retry on transient errors, distribution mode (even split or shared queue with first-come-first-served).
N out of M usernames added to your destination chat, with detailed per-account stats and failure breakdown.
First-touch DM campaigns from a playbook — personalised, paced, distributed
Sends first-touch direct messages to a target user list. Each campaign uses a playbook (intro template + follow-up sequence) with merge fields ({{first_name}}, {{username}}). Spreads sends across accounts and time windows, respects per-account daily caps and FloodWait, optionally adds the target to contacts before DM-ing to bypass privacy 'only contacts can DM me'.
N first-touch messages delivered with M replies, tracked per account, with replies routed to Neuro Chat for next-step handling.
Bulk DM dispatch to a known peer list with rate controls
Sends one or more message variants to a list of peers (users, channels, groups) the accounts already have a dialog with. Different from Cold Outreach: target list is people you've already engaged, the message is a broadcast not a first-touch. Useful for re-activation, announcement, or mass follow-up.
Bulk message delivered to N peers across M accounts within the configured cadence, with per-attempt audit log.
Extract member lists from channels and groups — bulk export of usernames, names, last-seen
Parses the member list of a target channel or group, extracts username, first/last name, last-seen status, premium flag, bot/scam/fake/deleted flags, optional phone (rare). Output is a paginated table you can filter and export to CSV. Used as the input for invite campaigns or cold outreach.
Filtered audience list of N members ready for export, sorting, or direct handoff to Invite / Cold Outreach.
Find Telegram channels by keyword, topic, or seed — feeds your targeting pipeline
Search Telegram for channels matching keywords, language, subscriber-count range, or 'similar to this seed channel'. Results include subscriber count, last-post date, language, ownership flags. Used to find where your target audience actually hangs out — the prerequisite to parsing them with Channel Members.
Curated list of N live channels matching your niche, ready to feed into Channel Sources or Channel Members.
Resolve phone numbers to Telegram accounts — find who's on Telegram from your existing list
Takes a list of phone numbers and resolves them to Telegram user IDs / usernames via Telegram's contact-import mechanism. Useful when you have CRM exports, lead lists, or any phone-based audience and need to know who has Telegram and how to reach them.
List of N phones with M resolved Telegram identities, ready to push into Cold Outreach or export.
Validate phone numbers and check Telegram presence — fast, batched, exportable
Takes a list of phone numbers and validates them in two passes: format check (correct international format, valid country prefix) and Telegram presence check (does this number have a Telegram account). Lighter-weight than Contact Discovery — doesn't resolve to a tg_id, just answers yes/no.
Cleaned phone list with Telegram-presence flag, ready to feed into Contact Discovery or to estimate campaign reach.
Drive signals — comments, reactions, stories, channel presence.
Make your accounts feel alive — and your posts feel popular.
Auto-comment under channel posts via your accounts — AI-generated, persona-rotated
Watches a target channel and posts contextual comments under each new post via accounts you've assigned. Comments are generated by an LLM from a persona prompt — each account has its own voice, so the thread doesn't read like a botnet. Includes latin-square first-commenter rotation, bimodal warmup (some accounts react fast, some hours later), per-post target counts, and FloodWait-aware pacing.
Each new channel post receives 3-7 humanlike comments + sub-replies within minutes-to-hours of publish, distributed across the account roster with persona-rotated voice.
Add reactions to posts across accounts — weighted-random emoji distribution
Adds reactions to a target post (or every new post in a channel) from your account pool. Each account picks an emoji from a weighted distribution you define — so you don't get 50 fire emojis in a row, you get a realistic mix that matches what real audiences leave.
N reactions appear on the target post over the configured window, with realistic emoji distribution and human-paced timing.
Publish Telegram Stories from your account fleet — bulk, scheduled, premium-aware
Schedules and publishes Telegram Stories from accounts in your fleet. Stories are Premium-only on the publish side — the panel auto-skips non-Premium accounts during bulk runs and surfaces the warning. Supports image, video, expiration period (6h/12h/24h/48h), per-account jitter so 50 stories don't go up at the exact same second.
N Stories live on N accounts within the chosen window, with view-tracking surfaced for the duration of the story.
Direct one-off engagement actions — react, view, join, leave
Lets you fire one-off engagement actions across your account pool: bulk-react to a specific post, mass-view a channel post, mass-join a group, mass-leave a chat. Different from Reactions Boost (one-shot, not a watching campaign) and Channel Sources (one action, not subscription tracking).
Action completed across the selected accounts within the jitter window, with per-account success/failure breakdown.
Track channels you care about — content feed for warmup, parsing, and AI-comments triggering
Maintains a curated list of channels the panel tracks: pulls new posts on a schedule, persists them so other features (Warmup, AI Comments, Reactions Boost) can read, react, and comment in realtime. The 'subscriptions list' for your accounts as a fleet.
Curated list of channels with up-to-date post history, ready to drive AI Comments, Reactions Boost, and realistic Warmup activity.
Manage Telegram channels owned by your accounts — create, post, pin to profile, delete
Each Telegram user can create up to 10 channels. This feature manages those channels at fleet scale: bulk-create channels per account, post content, pin to the account's profile (visible above their bio), unpin, delete. Used to make accounts look more 'real' (people who own a channel) and to set up account-bound content surfaces.
Each account in the fleet has 1-3 personal channels with credible content, optionally pinned to profile — significantly raising perceived legitimacy.
The infrastructure that makes everything else possible.
Keep accounts alive, credible, and recoverable.
Your fleet of Telegram user accounts — the foundation of every feature
Manages your pool of real Telegram user accounts. Each account is a logged-in session (encrypted at rest), bound to a country, optionally to a proxy and folder, with live status (active / spam-block / frozen / dead). Every other feature in the panel operates through these accounts — they're what actually sends messages, joins channels, and posts comments.
A live, healthy fleet of N accounts — geo-tagged, proxy-routed, status-tracked — ready to drive any feature.
Per-account network routing — SOCKS5, HTTP, MTProxy
Manages the proxy pool that accounts route their traffic through. Supports SOCKS5, HTTP, and MTProxy. Accounts can bind to a single dedicated proxy or to a rotating pool. The panel pings every proxy on a schedule and flags dead ones before they break a campaign.
Every account routes through a healthy, country-matched proxy — minimising cluster-detection risk.
Slow-drip account aging — read, react, subscribe like a real user
Simulates organic activity for fresh or dormant accounts. The worker schedules realistic actions across the day: open random channels, read posts, react to some of them, subscribe to a few. Spreads activity across hours and days so the account ages naturally before being used in campaigns.
Accounts go from 'just-bought looking' to 'normal user looking' in a week — measurably reducing first-action ban rate.
Recover spam-blocked accounts via @SpamBot — automated dialog, retry tracking
When an account hits spam_block, Telegram puts it in a 'limited' state where it can't message non-contacts. The classic recovery is to DM @SpamBot and answer its questions. Auto-Unfreeze does this automatically: opens the dialog, parses the bot's questions, picks safe responses, navigates the appeal flow, and reports back what happened.
Spam-blocked accounts attempt automated recovery; ~30-50% return to active state, the rest are flagged for manual review or retirement.
AI-generate unique personas for your account fleet — name, bio, avatar, username
Creates unique, internally-consistent personas and applies them to selected accounts: first/last name, bio, avatar (DALL-E), username. Each persona is generated from a brief (e.g. 'Russian-speaking female, 24-32, fitness/wellness, Moscow') and is unique across accounts in the same run.
N accounts now have unique, consistent identities matching your brief — name, bio, avatar, optional username.
Per-account ban risk dashboard — surfaces who's about to die before it happens
Computes a risk score (0-100) for every account in your fleet based on signals: recent FloodWait events, spam_block history, frozen-write probe results, action volume relative to age, proxy health, days-since-last-warmup. Surfaces the top-N at-risk accounts so you can intervene (cool down, warm up, retire) before they get banned.
Visibility into which accounts are heading for a ban, with enough lead time and breakdown to take preventive action.
Smart playbooks, A/B tests, voice replies, group dialogues.
Automate the work behind the work.
Multi-account scripted dialogue in a group chat — looks like a live conversation
Runs a scripted conversation between multiple accounts in a target group chat. You write the script (line by line, who says what), the worker delivers it with realistic timing (typing indicators, pauses, occasional 'human delay' if a real user posts). Used to seed group chats, simulate active discussion before organic users arrive, or run promotional dialogues that surface a product naturally.
Group chat appears actively used with a coherent multi-person discussion, optionally landing a product mention naturally inside the dialogue.
AI-powered DM lead handling — qualifies, replies, escalates to human when needed
Replies to incoming DMs across all your accounts via an LLM-driven playbook. Each playbook defines a goal (qualify lead, book a call, sell a product), qualifying questions, persona voice, and escalation rules. Conversations are visible operator-side — you can take over manually any time. Voice messages can be auto-replied with TTS using a per-playbook voice sample.
Inbound DMs handled at scale, leads triaged into qualified/disqualified, only the qualified ones escalated to a human — recovering operator time for high-leverage work.
Mass-report a target message or peer to Telegram abuse — for spam, scam, or violence
Submits abuse reports to Telegram from your account fleet against a target peer (user/channel/group) or a specific message. Each account picks a reason category (spam, scam, violence, fake, child abuse, illegal drugs, personal details, copyright). Per-tenant daily cap (default 100 reports/24h) prevents over-use.
N abuse reports submitted from N accounts against the target, with per-attempt audit log. Outcome (action / no action) shows in Telegram's notification to the reporting accounts within hours-to-days.
A/B test variants of any feature — content, pacing, persona — measure quality not vibes
Defines a test cycle: 2-N variants of a feature (e.g. two AI Comments persona prompts, three Cold Outreach intros, two pacing strategies), runs them in parallel for a fixed window, collects metrics (delivery rate, reply rate, ban rate, FloodWait rate), and surfaces a winner with confidence intervals.
Statistical answer to 'which variant works better' with confidence interval — replaces gut-feel with data.
Standalone AI utilities — batch text generation, persona drafting, message variant creation
A collection of one-shot AI utilities outside the main feature pipelines. Generate 50 message variants from one brief, draft 10 persona prompts for AI Comments, write 5 cold-outreach intros in different voices. Powered by Claude (Anthropic) — uses the panel's global key or your own BYO key.
Batch of N reviewed, formatted text artifacts (intros / personas / messages) ready to feed into the relevant feature.