Online task
Finishes a real job on a real website instead of handing you a link.
The test
- Day 1
Get me a table for 2 at restaurant this Saturday at 7:30pm. If 7:30 is gone, anything between 6:30 and 8:30 works. Use my name and number.
Setup
- One of five restaurants that take free-cancellation bookings on Resy or OpenTable, rotated so no restaurant gets two test bookings on the same night.
Time allowed: 30 minutes.Afterwards: Cancelled within 10 minutes of confirmation.
Passes when
- Books the named restaurant, for 2, between 6:30 and 8:30
- Uses the tester's own name and phone number
- Gets a confirmation by email or text and reports the time back
Scoring notes
- Only sends a booking link: 3.
- Books outside the time window: checkpoint 1 is missed.
Results 35 tested
19 passed, 6 partial, 10 failed. Average score 6.1.
| # | Assistant | Result | Score |
|---|---|---|---|
| 1 | OpenInstinct | Pass | 10 |
| 2 | Dots | Pass | 9 |
| 3 | Instinct | Pass | 9 |
| 4 | Ollie | Pass | 9 |
| 5 | Bantr | Pass | 8 |
| 6 | Caddy | Pass | 8 |
| 7 | Hark | Pass | 8 |
| 8 | Muse | Pass | 8 |
| 9 | Today | Pass | 8 |
| 10 | Asmi | Pass | 7 |
| 11 | Buoy | Pass | 7 |
| 12 | Familiar | Pass | 7 |
| 13 | Folk | Pass | 7 |
| 14 | Neo | Pass | 7 |
| 15 | OpenClaw | Pass | 7 |
| 16 | Orbits | Pass | 7 |
| 17 | Pally | Pass | 7 |
| 18 | szn | Pass | 7 |
| 19 | Tomo | Pass | 7 |
| 20 | Gemini Spark | Partial | 6 |
| 21 | Luke | Partial | 6 |
| 22 | Waldo | Partial | 6 |
| 23 | Del | Partial | 5 |
| 24 | Joshu | Partial | 5 |
| 25 | Persona | Partial | 5 |
| 26 | Ask Ranger | Fail | 4 |
| 27 | Lucas | Fail | 4 |
| 28 | Manus | Fail | 4 |
| 29 | Poke | Fail | 4 |
| 30 | tinyNature | Fail | 4 |
| 31 | Vellum | Fail | 4 |
| 32 | AVA | Fail | 3 |
| 33 | Brea | Fail | 2 |
| 34 | Tab | Fail | 2 |
| 35 | Wajo / Fo | Fail | 2 |
What people say 41 public quotes
Public posts about this kind of task, each linked to the original. Not scored, and posts by the companies themselves are left out.
People have reported that when they tried to disconnect the agent from their email, they found out it was retaining a copy of their inboxes anyway.
BugOpen sourceIt needed one more code before it got through. Then it reported that the membership was canceled and that there were no scheduled payments, and I received the confirmation email.
PraiseOpen sourceThe agent, which Meta launched on September 8, booked a Friday date night and made a packing schedule for my upcoming move. It also emailed a friend about a trip next month and made a planning document with places we want to visit.
Use caseOpen sourcealthough Muse almost could do it, it couldn't complete the final payment section without the site timing out. I ended up having to do it on my computer.
BugOpen sourceYou can use Instinct for just about everything: phone-only restaurants, dentist waitlists, cable bills.
PraiseOpen source@Muse is running my family's admin now: ordering school lunches for the week for my kids, booking swim lanes on a schedule as they open up, budget logging, the works. It's now saving me the $200/mo in other subs. Worth it.
PraiseOpen sourceThen it offered to create a Spotify playlist of those episodes so I don't have to scroll endlessly to find them.
PraiseOpen sourceInstinct should be provide address and contact information and go to https://ups.com @UPS website and complete everything to create a return shipping label for an international package.
Use caseOpen sourceI've used it every day for the last 7 days, and I'm highly, highly impressed.
PraiseOpen sourceI tried the same Airbnb queries that I did with Muse. It took literally an hour, I kid you not, for Instinct to respond to my aforementioned Airbnb query.
ComparisonOpen sourceToday I asked Muse to do it. It was very slow compared to my repetitive tapping, but it clipped every coupon for me after getting my login credentials for the grocery store's website.
Use caseOpen sourceIn a documented hands-on test, Muse shortlisted a Patagonia jacket at a discount that had already expired, apparently pulled from a batch-updated merchant catalog feed instead of live pricing.
BugOpen source@muse is already tightly integrated with link (native integration not just a skill) and it booked me tickets, while Instinct is still searching for the RSVP!
ComparisonOpen source@muse is already tightly integrated with link (native integration not just a skill) and it booked me tickets, while Instinct is still searching for the RSVP!
ComparisonOpen sourceTask depth. Muse opens browsers, fills forms, makes purchases, schedules lessons. Poke handles calendar, email, and scheduling. Muse's action scope is broader.
ComparisonOpen sourceIt books, it orders, it watches my inbox, it remembers everything.
PraiseOpen sourceAi can figure out where is my missing FedEx delivery and chase them! 🥳
Use caseOpen sourceMy Vellum assistant can now use my actual Chrome browser instead of a sandbox browser.
PraiseOpen sourceI regularly use pretty much every frontier model - ChatGPT, Claude, Gemini, etc. - but Instinct’s ability to actually navigate the web and get things done has been
ComparisonOpen sourceIn five days, Sheel Mohnot used Instinct to find an in-network podiatrist and fill in the forms, negotiate a Comcast bill from $100 to $60, book a DMV appointment, and link two separate United reservations.
Use caseOpen sourceInstinct found the local pub I wanted, dealt with its slightly idiosyncratic web form, and made the booking.
PraiseOpen sourceI trust it to run personal errands (e.g., book movie tickets), but not much beyond that.
Use caseOpen sourcealso books my movie tickets and holds the booking ids and seat numbers so i just show up. nothing on amazon or flights yet but for orders the hands-off part is real
PraiseOpen sourcealso books my movie tickets and holds the booking ids and seat numbers so i just show up. nothing on amazon or flights yet but for orders the hands-off part is real
PraiseOpen sourceI tried to get my agent to make a car park booking but because I couldn't use onelink (not available in the UK), I tried using my credit card details stored in the vault but that failed because it got stuck on a captcha
BugOpen sourceLoving it so far for job hunting. It is finding roles at companies I have never heard of that haven't popped up on LinkedIn, Indeed, etc.
PraiseOpen sourcewhat I like most is that it doesn’t feel like chatting with an AI, you ask for something, then there’s this natural little wait while it actually goes and tries to do it, and it comes back only when it has something useful
PraiseOpen sourceMagical experience booking a flight for my son and making dinner reservations. As a consumer offering, I think it’s the best AI product since ChatGPT launched in 2022.
PraiseOpen sourceBack came two options with times and prices, plus which one lets you skip the line by booking ahead.
PraiseOpen sourceInsane, someone used @wajo_ai ’s fo to send us donuts 🥹💗 Full circle ⭕️ Thank you #4some
PraiseOpen sourceI've tried Hermes, OpenClaw, Tasklet, GrokBot but Instinct takes the cake
ComparisonOpen sourceJust nuked my Instinct account. Tried it out for 48 hours and it does feel like the future of personal AI assistants (had it shop for new homeowners insurance, book a doctor's appointment online, search for travel, manage my LinkedIn notifications). I'm just... not yet
ComplaintOpen sourceHaving my AI assistant finding my tee time might be my favorite use case for Text Town
PraiseOpen sourcePally now communicate with my client in my own style and tone.
PraiseOpen sourceHowever, I think they have nailed computer use which is the hidden gem here.
PraiseOpen sourceBookings, rebookings, complicated travel planning, restaurant reservations, email follow ups, all out of the box, autonomous and done well and in a reassuring way (screenshots, etc).
PraiseOpen sourceI asked it to order me drunken noodles and BOOM! drunken noodles at my doorstep.
PraiseOpen source