Prime Air vs. Zipline vs. Wing: Who's Actually Delivering by Drone in 2026

In 2025, Zipline announced it had crossed 1.4 million cumulative commercial deliveries. That same quarter, Wing (Alphabet) reported roughly 450,000 deliveries since 2019, and Amazon Prime Air — after years of promises — was still operating in just two cities: College Station, Texas, and Tolleson, Arizona, with volumes the company prefers not to break down. Drone delivery has moved off the investor slide and into the field. But the 2026 scoreboard is far less even than the marketing suggests.

This comparison of the leading drone delivery companies in 2026 looks at all three by the numbers that actually matter to operators: daily flight density, aircraft architecture, retail integration, and cost per drop. No cheerleading.

What a Drone Delivery Service Actually Looks Like in 2026

A drone delivery service is a regular operation, approved by a civil aviation authority (FAA in the US, EASA in Europe, CAA in Rwanda, ANAC in Brazil), in which unmanned aircraft carry goods on BVLOS routes — beyond visual line of sight — between a hub and an end recipient, integrated with retail or healthcare ordering systems. In 2026, the three operators that matter at scale are Zipline, Wing, and Prime Air. Each is built on a radically different engineering thesis.

Photo: Max Nüstedt / Unsplash

Zipline: The Operator That Became Infrastructure

Zipline is the only one of the three that already behaves like public infrastructure in some markets. In Rwanda and Ghana, it has been moving blood, vaccines, and medicine for seven years. In the US, it operates with Walmart, Jet's Pizza, Sweetgreen, and the Cleveland Clinic hospital network.

The new platform — Platform 2, or P2 Zip — changes the game. Instead of landing, the drone hovers at 90 meters and lowers a droid on a tether: a self-stabilizing capsule that touches down in a yard the size of a dining table. That solves the problem Wing and Prime Air still face — delivering into dense residential neighborhoods without spooking the dog, the toddler, or the neighbor.

• Range per flight: about 10 usable miles on the P2 (round trip: ~20 miles).
• Payload: up to 3.6 kg (8 lb) on the P2, versus 1.75 kg on the previous system.
• Operational density: hubs in Dallas-Fort Worth, with Walmart, doing hundreds of deliveries per day per store in 2025.
• Stated cost target: $1 per delivery at full scale. Ambitious — but the fixed-wing plus droid architecture is the one that gets closest to it.

My read: Zipline is the only company that has solved the aerial last-mile elegantly. The droid is the most underrated piece of engineering in the sector.

Photo: Magic K / Pexels

Wing: Multirotor, Short Range, Walmart-Dependent

Wing came at it from another angle. The company was born inside Google's X Lab in 2012 and bet on a small hybrid tiltrotor optimized for landing and drop-offs across Australian suburbs. Logan, in Queensland, was its main laboratory. By 2023, Wing had already passed 350,000 deliveries across its various programs and markets.

In the US, the Walmart partnership that began in 2023–2024 in the Dallas-Fort Worth area brought real scale — and also exposed the limits of the architecture. The aircraft has a range of about 6 miles, a payload capacity of 1.2 kg, and needs to operate very close to the ground to complete a drop. In neighborhoods with tall trees, tight yards, or clutter around homes, that becomes an operational headache.

Wing's answer in 2025 was the Wing Delivery Network, a system that lets different aircraft fulfill orders from a shared network of operating nodes. The concept is sophisticated on the software and logistics side. But there's a ceiling technology can't lift: the aircraft's range is still relatively short, and payload stays capped.

Prime Air: The Promise That Keeps Slipping

Amazon promised drone delivery in 30 minutes back in 2013. It's now 2026. Let the record show that.

The Prime Air program has cycled through three aircraft. The MK27 never scaled. The MK30, approved by the FAA for BVLOS in May 2024, is the current bet — quieter, capable of flying in light rain, with a stated range of 15 km. Operations in College Station were suspended for three months in 2024 after two incidents with the MK30 during testing in Pendleton, Oregon. The service returned in November 2024 with expanded scope in Tolleson.

Prime Air's core problem isn't technical — it's organizational. Amazon wants to deliver items from the general catalog (average weight ~2.3 kg), which forces a large, loud, expensive aircraft. And the typical Amazon customer already gets free same-day delivery by truck. The value proposition of the aerial version? Frankly, nobody has explained it well.

• Active cities in 2026: 2 (Texas and Arizona).
• Eligible SKUs: a few thousand, out of a catalog of hundreds of millions.
• Estimated volume: low four digits per month per hub — numbers Zipline hits in a single good day at one Walmart store.

Photo: Joshua Brown / Pexels

The Comparison That Matters: Density, Cost, Architecture

Line the three up and the picture looks like this:

• Cumulative volume through end of 2025: Zipline ~1.4M | Wing ~450k | Prime Air <50k (estimated, not disclosed).
• Operational range: Zipline P2 ~16 km | Wing ~10 km | Prime Air MK30 ~15 km.
• Max payload: Zipline 3.6 kg | Prime Air ~2.3 kg | Wing 1.2 kg.
• Delivery method: Zipline (tethered droid) | Wing (low hover + winch) | Prime Air (vertical landing in a yard).
• Retail integration: Zipline (Walmart, Jet's, healthcare) | Wing (Walmart, DoorDash) | Prime Air (Amazon only).

Cost per delivery is the number nobody publishes honestly. Analyst estimates (ARK, McKinsey, Skyports reports) put Zipline in the $4–8 range today, on a path toward $1–2. Wing lands somewhere around $8–13. Prime Air sits above $30, and probably well above. It doesn't come close to sustainable unit economics.

And the Rest of the World? Meituan, Manna, and DroneUp

Worth noting: outside the Anglo trio, the game is different. Meituan, in China, crossed 450,000 deliveries in Shenzhen by early 2025 — hot food, vertical urban corridors approved by the CAAC, operations off shopping-mall rooftops. Ireland's Manna is delivering coffee and pharmacy items in Dublin at numbers that stand out for their density per km². And DroneUp, the American player, pivoted from direct delivery to outsourced infrastructure after losing its exclusive Walmart contract.

In Brazil, ANAC continues with RBAC-E 94 and a case-by-case BVLOS regime. iFood ran pilots with Speedbird in Aracaju and Campinas — short flights between fixed points, with landings at a base station. It's still a pilot. And honestly, it will stay a pilot until Brazil has a national Remote ID and a working UTM.

Photo: MART PRODUCTION / Pexels

Final Take

The 2026 scoreboard is not a three-way tie.Zipline is out front on publicly demonstrated scale and has arguably the most differentiated delivery architecture. Wing sits second with a defensible suburban thesis. Prime Air is chasing a value proposition that hasn't convinced even Amazon's own executives (see: the drone team layoffs in 2024).

If you're a retailer thinking about a pilot, talk to Zipline first — they're the only ones with a mature hub-in-a-box.

If you work in healthcare or critical logistics, Zipline has been the reference in emerging markets for seven years.

Keep an eye on Meituan. Within 24 months, it could have more cumulative deliveries than Wing and Prime Air combined.

And be skeptical of marketing numbers. Ask for "deliveries per hub per day," not "cumulative deliveries since 2019." That's where the conversation gets honest.