Keyword auto-replies
A rule matches a keyword and sends your template. That's the whole mechanism — which is why it's completely local: no AI is involved and nothing leaves your machine. Rules are unlimited on every plan.
Fully local · no AI
Auto reply
WhatsApp auto reply
Answer the questions you've already answered a hundred times — by keyword, entirely on your machine, or by what the customer actually meant. Every reply is held for a few seconds first, and typing cancels it.
7-day free trial · no credit card · works with your existing number
Two kinds
Match the words, or understand the question
Most of your inbox is the same five questions. One kind of auto reply matches how they're worded; the other works out what was asked. They're built differently, and it matters where your data goes.
Semantic auto-replies
“Do you have a minimum order?” and “what's the MOQ?” are the same question, and a keyword list will miss one of them. A semantic reply answers both — from the knowledge base you wrote in your own words: prices, policies, the FAQs you keep retyping. It answers from those facts; it doesn't invent them.
AI · opt-in · Professional & Expert
The grace window
It waits, where you can see it
This is the part that makes an auto reply survivable: nothing is sent the instant it matches.
Held for a few seconds, visibly
When a reply is about to go out it's held first, with a countdown above your composer. You can see what's queued and how long you have — it isn't happening somewhere you can't watch.
Start typing and it's cancelled
Type in that chat, or send anything yourself, and the pending reply is dropped — you've taken over. You never have to race the automation for your own conversation.
It doesn't guess whether you're there
There's no presence detection here, deliberately. No heuristic deciding whether you're at your desk — that kind of guess is wrong at exactly the wrong moment. You get a window you can always grab instead.
Beyond the window: replies only fire under the conditions you configure, a per-contact cooldown stops one person being answered over and over, and a daily cap limits the total that can go out.
How it works
Three steps, inside WhatsApp Web
Write the answer once
Save the reply as a template and point a keyword rule at it. For semantic replies, write the facts — prices, MOQ, lead times, policies — into your knowledge base as plain text.
Say when it may fire
Set the conditions a reply has to meet, the per-contact cooldown and the daily cap. AI stays off entirely until you switch it on.
Watch the first ones land
When a reply matches, the countdown appears above your composer. Let it run and it sends; type and it's yours again. Everything happens in the WhatsApp Web you already use.
Read this part
What an auto reply can't do for you
It buys you minutes at 2am. It doesn't buy you a business that answers itself, and we'd rather say so than sell you one.
It's not an unattended bot
And it shouldn't be sold as one. It covers the questions you've already answered a hundred times, at hours you're asleep. It doesn't replace you — anything it can't answer from your knowledge base should reach you rather than be improvised.
Automating WhatsApp Web carries risk
It isn't officially supported. That's real terms-of-service risk, and no tool — ours included — can remove it. The window and the cooldown keep the automation something you're steering; they aren't a promise about your account, and anyone offering you immunity is selling you a feeling.
What the AI sees, precisely
AI is off until you turn it on. When it's on, the extension discloses what was sent and to which provider, and one request carries at most the 40 most recent messages or 24,000 characters of that chat. Wanova never stores your chat contents. Keyword rules send nothing at all.
Our advice, against our own sales interest: automate the questions whose answers never change. If an answer is one you'd have to honour later — a price, a delivery date, an exception — let it reach you instead. A reply you didn't write is still one you'll be held to.
And the rest?
The questions it hands back to you
An auto reply covers the answers you already know by heart. The conversations that decide a deal are still yours — which is why Wanova is a co-pilot first.
Still thinking about it — what's your best price?
I can hold $4.80/unit at 500 pcs through Friday.
You always press send
Send
When you'd rather approve each one
For the replies that actually matter, the co-pilot writes the draft and drops it in your composer. You edit it, you send it.
You always press send
¿Cuál es el precio para 500 unidades?
What's the price for 500 units?
Sure — $4.80 per unit at 500 pcs.
Claro — $4.80 por unidad a 500 pzas.
Answering across a language line
Incoming messages are translated as they land, and you reply in your own language while Wanova sends theirs.
Opt-in · cachedNeeds you 3
AKAhmed K.4d silent
MRMaria R.2h waiting
CWChen W.promised
What it escalates to you
The ones no auto reply should touch: who's still waiting, who went quiet after a quote, what you promised and when.
Runs locally
Plans
Auto replies are on every plan
One price per account — not per person and not per number. Every plan covers several WhatsApp numbers and as many teammates as share them. The AI co-pilot starts on Professional.
Starter$14.99/moup to 3 numbers · every local tool · 3,000 AI credits
Professional$24.99/moup to 10 numbers · AI co-pilot · 5,000 AI credits
Expert$34.99/mounlimited numbers · 7,000 AI credits · webhooks
Keyword rules run locally on every plan · See full pricing
FAQ
Questions, answered
Can I set up WhatsApp auto replies without AI?
Yes — keyword auto-replies are exactly that. A rule matches a keyword and sends your template, and it runs entirely on your machine: no AI, and nothing leaves it. Every plan includes them, and there's no cap on how many rules you create.
What's the difference between keyword and semantic auto replies?
A keyword rule fires on the words you list. A semantic reply understands what was meant — “do you have a minimum order?” and “what's the MOQ?” get the same answer — and writes it from your own knowledge base rather than inventing it. Semantic replies are AI, opt-in, and available on Professional and Expert.
Can I stop an auto reply before it sends?
That's the grace window. Every auto reply is held for a few seconds first, with a countdown above your composer. Start typing in that chat, or send anything yourself, and the pending reply is cancelled — you've taken over.
Does it know when I'm away?
No, and that's deliberate. Wanova does not try to detect whether you're present — there are no heuristics guessing whether you're at your desk. Instead every reply pauses in front of you and typing cancels it, so nothing has to guess right.
Will it make up an answer it doesn't have?
It answers from the knowledge base you wrote — your prices, policies and FAQs. It is not an unattended bot: anything it can't answer from what you gave it should reach you rather than be improvised. It buys you minutes at 2am; it doesn't replace you.
What does the AI see, and where does it go?
AI is off until you switch it on. When it's on, the extension discloses what was sent and to which provider, and a single request carries at most the 40 most recent messages or 24,000 characters of the current chat. Wanova never stores your chat contents.
Is auto replying on WhatsApp risky?
Automating WhatsApp Web isn't officially supported, so it carries real terms-of-service risk that no tool can remove — ours included. The grace window, the per-contact cooldown and the daily cap exist so the automation stays something you're steering. They are not a promise about your account.
Related
Keep reading
Answer the easy ones. Keep the ones that matter.
Try every feature free for 7 days. No credit card.
Start free trial