Your website, on the phone.
Loquin is an AI assistant for websites. It reads everything you’ve published and answers questions about it — in a chat widget, out loud in the browser, and on a real phone number that rings 24 hours a day. One line of script. No plugin. The core is open source and free to self-host.
Self-host the core for free, forever. Pay only for the voice we run.
<!-- paste before </head> -->
<script src="https://cdn.loquin.io/v1.js" data-site="lq_live_a1b2c3" defer></script>
Three channels, one brain
The same answers, however people reach you.
Most tools give you a chat box. The people most likely to buy from you are still picking up the phone — and hanging up when nobody answers. Loquin indexes your content once and serves all three surfaces from it, so the phone never contradicts the website.
Chat
A grounded chat widget that answers from your pages, posts and products — and says “I don’t know” when the answer isn’t on your site, instead of inventing one.
Browser voice
Visitors press a button and talk. Sub-second responses, because the index sits next to the model rather than on your web host.
Phone
Your own number, answered on the first ring, at 2am, from the same knowledge. Takes messages, captures leads, hands off to a human when it should.
Who it’s for
Built for businesses where a missed question costs money.
Loquin works on any website, but it earns its keep where people still phone up and where the answer is already published somewhere on your site.
Clinics and practices
Opening hours, what you treat, whether you take a given insurer, how to prepare for an appointment. The questions that fill a receptionist’s morning are almost all already answered on your site.
Trades and home services
A missed call is a lost job. Loquin picks up on the first ring at 7pm, confirms you cover the caller’s area, and takes the details while the lead is still warm.
Legal and professional services
Answer practice-area and process questions from your own published content, capture the matter details, and route anything that needs a qualified person to a qualified person.
Ecommerce and WooCommerce
Shipping, returns, sizing, stock, and “do you have this in blue”. Product data comes straight from your catalogue rather than a stale copy of it.
Membership and course sites
Curriculum questions, access problems, billing and renewals — handled from your documentation instead of your inbox.
Agencies and studios
Deploy across a client portfolio from one dashboard, white-labelled as your own. One script tag per site, one place to manage all of them.
How it works
Paste one line. Loquin does the reading.
Drop the script in your header
Any theme, any page builder, any host — or a header-scripts field if you have one. Nothing to install, nothing to update, no PHP to break on your next WordPress release.
We read your site properly
On WordPress we pull structured content straight from the REST API — real titles, real bodies, real taxonomies, including custom post types and WooCommerce products. If that’s locked down we fall back to your sitemap, then to crawling. Most tools only ever crawl, and it shows in the answers.
It stays current
Publish an edit and the index follows. No re-training, no re-uploading, no forgetting you changed your prices in March.
Turn on voice when you want it
Same knowledge base, two more channels. Claim a phone number in the dashboard and point your existing line at it, or use it as an after-hours overflow.
Open source
The core is yours. Run it yourself, forever.
Loquin Core is open source and self-hostable. Ingestion, chunking, embeddings, retrieval and the chat widget — the whole pipeline, on your own infrastructure, with your own API keys. You pay your model provider directly and we take nothing.
We don’t cripple it to sell you something. There’s no locked feature, no trial clock, no seat limit. If self-hosted chat is all you need, that’s a complete product and it’s free.
At a glance
The short, factual version.
Everything you’d want to know before deciding whether to read the rest.
| What it is | An AI assistant that answers questions about a website using that website’s own published content |
|---|---|
| How you install it | One <script> tag in the page header. No plugin, no build step, no framework |
| Works with | Any website. WordPress, Webflow, Shopify, Squarespace, static sites, custom apps |
| Channels | Text chat, in-browser voice, and inbound phone calls — all from one knowledge base |
| How it reads your site | WordPress REST API first, then sitemap, then crawl. Post types are discovered, not assumed |
| Open source | Loquin Core, MIT licence, fully self-hostable with your own API keys |
| Free tier | Unlimited chat conversations and unlimited indexed pages when self-hosted |
| Paid plans from | $249/month, billed annually, including 500 voice minutes |
| Voice billing | Plain minutes. Not credits, not tokens, not message units |
| Model providers | Bring your own OpenAI or Anthropic key on Core; managed on hosted plans |
| Training on your data | Never. Your content is not used to train any model |
| Made by | Cenango, a US software company building software since 2003 |
Pricing
Free where you run it. Metered where we do.
Voice is billed in plain minutes. Not credits, not tokens, not “message units” that cost between two and six of themselves depending on the weather.
Core
- Unlimited chat conversations
- Unlimited indexed pages
- Bring your own model keys
- REST, sitemap and crawl ingestion
- Full source, no locked features
- Community support
Voice
- Everything in Core, fully hosted, CRM
- Browser voice + one phone number
- 500 voice minutes included
- Managed index, always current
- Transcripts and call summaries
- Lead capture and routing
- Transfer to a human
- Overage at $0.35/min
Agency
- 5 client sites
- 3,000 pooled voice minutes
- White-label widget and dashboard
- Client sub-accounts
- Set your own client pricing
Questions
The things people actually ask.
What is Loquin?
Loquin is an AI assistant for websites. It indexes the pages, posts and products you’ve already published, then answers visitor questions from that content across three channels: a chat widget, voice in the browser, and a real inbound phone number. You add it with a single script tag. The core is open source under the MIT licence and can be self-hosted for free; paid plans add the voice channels, which need infrastructure we run.
How much does Loquin cost?
Loquin Core is free and always will be — it’s open source, self-hosted, and has no usage caps or locked features. Hosted plans start at $249 per month billed annually, which includes browser voice, one phone number and 500 voice minutes. Voice Pro is $499 for 1,500 minutes, and the Agency plan is $999 for five client sites and 3,000 pooled minutes. Overage is $0.35 per minute.
How is Loquin different from Chatbase, Intercom or Tidio?
Three practical differences. Loquin gives you a real phone number answered from the same knowledge base, not just a website widget. The core is genuinely open source and self-hostable, so you can run it without paying us anything. And on WordPress it reads structured content from the REST API — including custom post types and WooCommerce products — rather than only crawling your HTML, which produces noticeably better answers.
Is this WordPress-only?
No. It’s a script tag, so it runs on anything — Webflow, Shopify, Squarespace, a static site, a custom app. WordPress just gets better ingestion, because we can read the REST API instead of guessing at your HTML.
What stops it making things up?
Retrieval happens before generation, every time, and the model is instructed to answer only from what it retrieved. When nothing relevant comes back it says so and offers to pass you a message rather than improvising. In voice this matters more than in chat — a chat widget can show you a source link, a voice agent has to be right out loud.
Where does my content go?
On Core, nowhere — it never leaves your infrastructure. On hosted plans your indexed content sits in your own namespace on our servers, and we don’t train models on it. You can delete it and it’s gone.
Do I need a new phone number?
You get one, but you don’t have to publish it. Most people forward their existing line — either always, or only after hours and when it’s busy. Callers hear that they’re speaking to an automated assistant at the start of the call.
What happens when it can’t help?
It takes a message and sends it to you, or transfers to a human on Voice Pro. An assistant that knows its limits and hands over cleanly beats one that keeps talking — and callers can tell the difference.
Why is voice priced so much higher than chat tools?
Two reasons, and we’d rather say them plainly. Voice genuinely costs real money per minute to run. And cheap AI subscriptions have terrible retention because nobody sticks with a $29 tool — we’d rather charge enough to support the product properly and keep it running than race to the bottom and disappear.
Can I switch off the Loquin branding?
On Core it’s off by default — it’s your site. On the Agency plan the whole thing white-labels, including the dashboard your clients log into.
Early access
Be one of the first sites that can talk back.
We’re onboarding a small first group by hand so we can tune each voice properly. Tell us about your site and what you’d want it to handle.
Prefer to just read the code? Loquin Core is on GitHub.