Update, October 2026: since 29 September 2026 Grazi runs on Claude Sonnet 5.5. The post below describes the July move to Sonnet 5.
Grazi has one job: help you get an answer and get on with your day. Not another dashboard to learn, not a chatbot that reads you the manual back. You ask a plain question, it looks at your actual business, and you keep moving.
We work on Grazi every single week. We read the answers it nailed and the ones it didn't, see what it got right, and fix what it got wrong. It's a real routine, not set-and-forget, and bit by bit it's been getting sharper.
This update isn't bit by bit. We've moved Grazi onto Claude Sonnet 5, Anthropic's newest model at the time, and it's the biggest single jump in how capable Grazi has been since we launched it. Here's what actually changes for you.
It understands the messy question, not just the tidy one
Real questions on a busy day aren't neatly worded. "Why did that refund end up on the wrong card and how do I stop it happening again?" is two questions and a half-remembered detail rolled into one, fired off between clients.
Sonnet 5 is much better at untangling that. Grazi follows the thread, works out what you actually meant, and answers the real question instead of giving a tidy answer to one you didn't ask. Fewer "that's not what I meant" moments, fewer rephrases, less back-and-forth before you're moving again.
The insights actually help
Most in-app AI hands you advice you could have found on Google. "Engage your customers." "Consider a loyalty programme." Thanks for that.
Grazi is plugged into your live data: your bookings, revenue, customers, pets and team. Ask "How am I doing this month?" and you get your numbers and your trend, not a definition of revenue. Ask "Which services are most profitable?" and it works out revenue per minute across your service types, which is the number that actually tells you where to spend your day. Ask "Why is Tuesday afternoon always empty?" and it reads your real booking patterns instead of guessing.
Sonnet 5 makes that second step, the reasoning on top of your data, noticeably stronger. It's the difference between handing you a number and telling you what you'd look at next.
Show it, don't describe it
You can drop a screenshot straight into Grazi. A Stripe receipt, a booking that looks wrong, a report you don't quite follow. Sonnet 5 reads what's in the image far better, so instead of typing three paragraphs describing what you're staring at, you show Grazi and get the explanation back in seconds.
Get more done in the same window
"Draft a message to clients with expiring vaccinations." Grazi pulls the real customers and pets, writes something that sounds like you, and you send it. The upgrade lands those drafts closer to right the first time, so there's less editing before they go out.
That's the whole point of the jump. Not to be clever, to be quick. Ask, get a straight answer, keep going.
Grazi doesn't replace us
Worth being clear on this one. Grazi is here so you can get unstuck at 9pm without waiting on anyone. It's not here to put a wall between you and our team.
When it can't fully sort something, one tap sends the whole conversation to us, context and all, so you're not repeating yourself to someone starting from scratch. And a thumbs up or thumbs down on any answer comes straight back to us and feeds the weekly work. You're not shouting into a void, you're shaping the thing you use every day.
Where this is heading
Sonnet 5 is a big step, and it is a step, not the finish line. We'll keep tuning Grazi every week, and we're building toward it handling more of the busywork for you, not just answering when asked. You can read where that's going in our piece on the future of pet business software.
It's live in your account today. Hit Cmd+K (or Ctrl+K on Windows), or open Grazi from any page, and ask it something real. "How am I doing this month?" is a good place to start.