MirageCHAT questions, answered clearly

MirageCHAT provides guided connection flows for eligible accounts. You do not build a separate Meta app, but platform account eligibility, requested permissions and provider review still apply.

Workspace creation takes minutes. Channel activation time depends on account eligibility and provider authorization; no server setup or customer-owned Meta app is required.

Any business that talks to its clients online: restaurants, beauty salons, gyms, shops, clinics, car rental, hotels, real-estate and travel agencies — and more. If your clients write to you on WhatsApp or Instagram, or you post on social media, MirageCHAT does the heavy lifting for you. There are pages with the details per trade: restaurants, clinics, car rental, salons, gyms and shops.

Eligibility depends on the number and Meta setup. MirageCHAT guides the connection without asking the customer to build a separate Meta app.

Yes — 5 days free, no credit card required. Cancel anytime.

WhatsApp, Instagram, Telegram and website chat are available now. Facebook Messenger and SMS require the relevant provider access and account setup. The Channels screen shows the exact abilities active for each connected account.

MirageCHAT supports media, location, contact cards and interactive formats where the connected WhatsApp or Instagram account and platform permissions allow them. The channel card shows the abilities available to that account.

To start a conversation, or to message someone more than 24 hours after their last reply, WhatsApp requires an approved template. You create templates in MirageCHAT, we submit them to Meta, and you are notified automatically when each is approved or rejected. Templates with a media header (for example a photo) are supported.

Eligible Instagram accounts can use the supported reply, hide, delete and private-reply actions when Meta grants the required permissions. MirageCHAT checks the account's abilities before offering an action.

A QR code that opens a WhatsApp chat with your number and a message already typed in — perfect for flyers, table tents, receipts, packaging and door signs. Create and manage them in Settings or via the API.

Yes. Every person who writes to you gets a contact record — which channel they came from, when they first wrote, every conversation since, and any notes your team adds. Open it from the top of any chat. Nothing is imported from anywhere else; it builds itself from the messages you already receive.

Yes — segment your customers by channel, tag or location, schedule a send, and track delivery. Opt-outs are always respected.

Yes. MirageCHAT is available on Google Play and the App Store. The mobile apps use the same workspace, connected accounts and permissions as the web app.

No. It is the same agent account. If someone can use the inbox on the web, they can use it on their phone — and you can turn mobile access off for any individual agent under Settings → Agents.

Yes. A reply typed straight into the notification is sent immediately, whether the app is open, in the background or fully closed. The notification turns into a tick once the reply is on its way.

Nothing is lost. The reply is kept on the phone and sends itself the moment the connection returns — even if the app has been closed in the meantime.

Most updates arrive on their own and install at a safe moment — never while somebody is typing. Now and then a release changes the app more deeply and a fresh download is needed; the app says so when that happens.

MirageCHAT is available on Google Play and the App Store. An Android download may also be provided from our site for supported installation scenarios.

The app has its own lock — fingerprint, face or a four-digit PIN — separate from the phone's own lock, so an unlocked phone left on a desk still does not open your inbox. You can also switch off mobile access for any individual agent, and their sessions stop working.

Google Play and the App Store currently list MirageCHAT 1.3.2. The apps can also receive compatible in-app updates; a new store build is required when native phone capabilities change.

Claude drafts replies in the customer's own language for availability, pricing and FAQs. You choose, per channel, whether the AI sends automatically or just suggests a draft for your team to approve.

From a knowledge base you write — prices, policies, opening hours, anything you would otherwise repeat fifty times a week. You can paste text or point it at a page on your website. The AI answers from that, and when the answer is not there it says so instead of inventing one.

The AI replies in the customer's own language — English, Italian, German, Spanish, French and more. It detects the language of the incoming message and answers in it, so your team writes once and every customer reads it in their own.

Yes. With Auto-Replies you can greet every new chat or send an away message outside business hours — automatically. You choose the channel and the exact accounts it applies to, and newly-connected channels appear on their own.

An auto-reply is one fixed message on one trigger — a first message, or a message outside business hours. A workflow is a sequence: ask the AI to answer, wait ten minutes, and if the customer is still unanswered, hand it to a person. Auto-replies are set up in a minute; workflows are built on a canvas when one step is not enough.

Yes. Turn on suggestion mode and the AI drafts a reply for your agent to review and send with one click, instead of answering on its own.

Yes. Social Auto Pilot schedules recurring Instagram and Facebook content from your media library. Instant Post is the separate publish-now workflow for selected accounts. TikTok customer activation depends on TikTok approval and account permissions.

Instagram and Facebook publishing are live for eligible connected accounts. The TikTok integration is built, but customer availability depends on TikTok approval and the permissions returned for the connected account.

Instant Post does not add music when it combines clips into a reel. Social Auto Pilot uses only media and audio that the workspace is entitled to use; platform music libraries remain governed by each platform and account.

Yes. Response time is measured from the customer's last message to your reply — the wait they actually experienced, not when someone opened the chat. You get it per agent, per channel and per day, with the AI listed alongside your team so you can see how much it handled on its own.

Yes — every reply carries the person who sent it and the device they used, phone or browser. Team Activity shows the whole team, including anyone who replied to nothing that week, which is usually the number a manager is looking for.

The audit log records sign-ins, permission changes, connected channels, AI prompt edits and every data export, each with the person, the device and the time. Entries are sealed with a fingerprint of the one before them, so altering or deleting one breaks the chain and a single button shows you exactly which entry it was. It cannot make tampering impossible — nothing stored in the same database can — but it makes tampering impossible to do quietly, which is what makes a log worth having.

Yes, with the Team Announcements module. Shift notes, urgent updates or team news arrive as a push notification on your agents' phones. It reaches your own team only — never customers, never another business — and every announcement is recorded in the audit log with who sent it.

Yes. An admin attaches a photo to any quick reply in Settings — directions to your door, the menu, a photo of the thing everyone asks about. When an agent picks that reply in a conversation, the photo and the text are sent together in one tap, from the web inbox or the mobile app. Saved replies are managed by admins, so the whole team sends the same approved answer.

Incoming messages can be translated on demand. The AI can draft a reply in the client's language from the conversation and your knowledge base; in suggestion mode an agent reviews and sends it.

Yes — that is what Alert Audit does. You add the words that matter to you (a competitor, for example) or switch on patterns like a phone number or a link to another site, and you are told when an agent reply contains one. If you do not know what to watch for, AI can read your own conversations once and suggest names it found — you decide which to keep. Alerts arrive as a single digest each morning and open the conversation for you to read. Nothing is ever closed, blocked or auto-corrected, and agents never see the list.

They are the answers your team sends every day, written once by an admin and inserted in one click. A quick reply can carry a photo — the map to your door, the menu, the one everyone asks about — which sends immediately with the text as its caption. It can carry links that land as their own tidy lines, and fields like the customer's name, which fills itself from the conversation, or the time and price, which stay as blanks the agent fills before sending. Replies are grouped into categories and searchable, so the right one is found in seconds even when there are dozens. You can also tie a reply to a single channel, so an Instagram link never shows up in a WhatsApp chat.

Yes. The Pro plan includes multi-location rollup and unlimited team members.

Yes. Your whole team works from one shared inbox — assign conversations, snooze, tag and add private notes, and see who is replying in real time.

Simple monthly plans in US dollars, starting with a 5-day free trial and no credit card required. Upgrade, downgrade or cancel anytime. Message-based channels like SMS and WhatsApp may incur carrier fees on your own provider account.

No. Meta charges per conversation at its own published rates, and that charge goes to your own WhatsApp Business account — it does not pass through us, so there is nothing for us to mark up. What you pay MirageCHAT is the package you chose, and the AI replies beyond what it includes. That is all.

Plans and optional modules determine what is enabled for a workspace. The exact active modules, limits and price are shown before confirmation; provider charges such as WhatsApp or SMS remain on the customer's provider account.

Your data is isolated per workspace and hosted in the EU. See our Privacy Policy for details.

Yes. When a customer replies STOP they are opted out automatically (START opts them back in), and opted-out contacts are never sent templates or campaigns. It is GDPR-friendly by default, and you can manage the list via the API.

MirageCHAT provides scoped API keys, supported messaging and channel-management endpoints, and webhooks. The exact public contract is in the authenticated developer manual; reseller use depends on the plan and commercial agreement.

Yes — that is what Integrations are for. If you have an API, connect it and map your own field names onto ours; the answer appears as a panel inside the conversation. If you already have a working screen — a planner, a booking page, a dashboard — you do not need an API at all: give us its address and it opens in a tab with the customer's details already filled in. You choose which accounts and which agents can see it.

Yes. Every conversation in the list shows the date and time of its last message - Today 14:32, Yest. 12:05, or 12 Aug 18:22 for anything older - rather than a rough 3h, so you can tell 14:32 from 14:59 when you are deciding what to answer first. Conversations still waiting for a reply keep their waiting badge in that corner instead. The clock follows your own device, and the browser and the mobile app read the same.