Product

DoorDash's AI Agent Takes Food Orders Over Text

DoorDash launched an AI agent that lets US customers order food by texting through Apple Messages, with a waitlist open now and drone tests on the way.

Daniel HarrisDaniel Harris
Heat: 900
A minimalist illustration of texting bubbles rising from a phone with a burger and takeout bag, warm restrained tones

DoorDash now lets you order food by texting an AI agent in Apple Messages. Say "order my usual" and it figures out the rest, no app required. A US waitlist is open now.

DoorDash's text-to-order AI agent, explained

DoorDash announced an AI agent that takes food orders over text. The catch that makes it interesting: you don't open the DoorDash app. You text it through Apple Messages, the same app you use to message friends.

You can send loose instructions like "order my usual," and DoorDash says the agent understands you mean your recurring order, the Friday-night standby rather than a literal request. You can also name a specific dish, ask for a local recommendation, or describe what you're in the mood for. The agent searches nearby restaurants, builds a cart from what you said, and can even text you photos of the dishes it suggests.

It reads like a small shift, but it's a real change in how ordering works. Instead of browsing a menu grid and tapping through options, you describe what you want in a sentence and let the agent assemble the order. DoorDash is opening a waitlist for US customers who want to try it.

How ordering food through Apple Messages works

Here's the flow. You message the DoorDash agent inside Apple Messages, the way you'd text a person. You type something conversational: "get me something spicy from that Thai place" or "order my usual." The agent interprets it, queries local options, and proposes a cart. You confirm, and it moves to checkout.

The group feature is the part most people will actually test. DoorDash says the agent can handle mixed dietary preferences and different quantities in a single order. So if half your team wants vegetarian and half wants chicken, you can say so once and get one cart instead of five separate orders. Anyone who's coordinated a group lunch knows how much that saves in back-and-forth.

The photos are a nice touch, and a useful one. Rather than asking you to imagine a dish from a menu description, the agent can send a picture. That's closer to how people actually decide what to eat.

Why DoorDash is betting on a personal AI agent

The launch is a direct play against Uber Eats and Grubhub, but the bigger context is a shift across tech. Companies are racing to build personal AI agents that handle tasks without making you navigate an app or a website. If an agent can order your food, it can plausibly book your table, reorder your groceries or handle your returns.

DoorDash also used the moment to announce it will begin testing delivery drones with select restaurants in Northern California. That's separate from the ordering agent, but both point the same way: the company wants to own more of the journey from craving to doorstep, and to remove friction wherever it can.

There's a strategic logic here worth naming. An app makes you come to it. An agent that lives in your messages comes to you. That's a harder habit to break, and it keeps DoorDash in the conversation even when you're not thinking about delivery.

What the DoorDash text agent means for users

If you order delivery often, this could trim real time off the process. The group-order handling alone is worth a look for anyone who regularly feeds a crowd with different tastes. If you're casual about delivery, the benefit is smaller, but the direction is still telling: the app you open may matter less than the assistant you text.

Two caveats before you get excited. The feature is waitlisted in the US, so availability is gradual and uneven. And an AI agent ordering food on your behalf raises a familiar question: what happens when it gets the order wrong? DoorDash hasn't detailed how refunds or disputes work when the agent, not you, made the call.

For now, the practical move is to join the waitlist if it appeals, and treat it as a beta. Test it on a low-stakes order before you trust it with a big group meal. Small risks first.

What to watch after DoorDash's launch

Keep an eye on how fast DoorDash expands beyond the waitlist and whether the agent's recommendations feel personal or generic. The real test isn't whether it can parse "order my usual." It's whether it learns your actual preferences well enough to order well on days you don't want to think about it. That's a higher bar than it sounds.

Share This Story

Sources

Related AI News

A search bar with an assistant sidebar sliding in from the left and shortcut chips underneath
Product

Google Put Your Chat History on the Search Homepage

The box in the middle of google.com still takes a query. What changed is the furniture around it, and the reasoning behind the change is about how many questions you're supposed to ask in a row.

Heat: 1,790
An outlined app window in the foreground with a second window working behind it and a cursor path running between them
Product

Gemini Can Now Drive Your Mac, in a Window You Don't Have to Watch

Google's Mac app can now take a multistep job and finish it in apps it was never connected to, inside a window you don't have to watch. The part worth reading is where it stops and asks.

Heat: 1,740
Chat bubbles sealed inside a vault door with a retention dial and two faded 72-hour timers
Product

Gemini Now Lets You Keep Your Chats and Still Say No to AI Training

One switch used to decide two things at once, and now it decides one. The second change runs the other way and keeps more of what you assumed was gone.

Heat: 1,500
A stack of library layers on a plinth with a locked version tag on top and streams feeding in from above
Product

llama.cpp Reaches v1.0.0 and Promises to Stop Breaking Your Build

llama.cpp reached v1.0.0 after three and a half years, and the release is about a promise rather than a feature. The promise is that the next breaking change to the C API gets a number before it breaks anybody's build.

Heat: 1,200