Starship Delivery Robot Review: Specs & Verdict

Starship Delivery Robot: Quick Buyer Verdict

The Starship Delivery Robot should be evaluated as a complete autonomous last-mile delivery system, not simply as a six-wheeled robot.

The current Starship e-model travels mainly on sidewalks and pedestrian routes, operates at Level 4 autonomy, carries approximately three shopping bags and is designed for local deliveries from stores to customers. Starship states that its fleet is more than 99% autonomous, with human remote assistants available when a robot encounters a situation that requires help.

The hardware is only one part of the offer. Starship also provides mapping, charging infrastructure, APIs, app integration, fleet supervision, maintenance, operations support and customer-service infrastructure.

That makes Starship substantially more deployment-ready than buying a robot chassis and building the delivery operation yourself. The trade-off is reduced control: organisations wanting to own the hardware, rewrite its autonomy stack or deploy it wherever they choose are looking for a different product.

Pros

• One of the largest and most commercially proven autonomous-delivery fleets in the world.
• Level 4 operation with more than 99% autonomy in deployed service areas.
• Current e-model is rated for up to 18 hours of operation.
• Approximately three shopping bags of insulated cargo capacity.
• 12-camera sensor suite including time-of-flight cameras, plus radar and ultrasonic sensing.
• Six-wheel chassis with a bogie system for curbs and uneven sidewalk transitions.
• All-weather operating experience including snow.
• Locked cargo compartment, tamper alarm and app-based customer unlocking.
• APIs for delivery jobs, status updates and cancellations.
• Can integrate into a retailer’s existing app or use a Starship-based customer application.
• Starship handles mapping, charging infrastructure, fleet management and maintenance.
• Strong real-world grocery evidence through S Group in Finland.

Cons

• No public per-robot purchase price or conventional standalone hardware-buying path.
• Usually requires a Starship-managed deployment rather than customer-controlled robotics development.
• Short operating radius compared with road vehicles.
• Approximately three shopping bags is insufficient for large grocery baskets.
• Published industrial payload is only 10 kg.
• Requires suitable sidewalks, crossings and mapped operating areas.
• Store staff still need to pick, pack and load customer orders.
• Does not solve autonomous movement through stairs or most building interiors.
• Economics depend heavily on order density and robot utilisation.
• Public-space deployments require local accessibility, operational and regulatory review.
• Buyers have less ownership of the autonomy stack than with an open robotics platform.
• Starship is winding down its U.S. university-campus operations in 2026.

Our recommendation: shortlist Starship if you are a grocery retailer, food-delivery platform or large site with a high volume of repeatable short-distance trips inside a geographically dense operating area. Do not begin with the robot specification. Begin with your order density, delivery radius, basket size and cost per existing delivery, then ask Starship to prove the economics through a controlled deployment.

Review the Starship Delivery Robot listing at Anton Robots and the Starship Technologies manufacturer page before comparing alternatives.

How Much Does the Starship Delivery Robot Cost in 2026?

Starship does not publish a standard retail purchase price for its current delivery robot.

This is one of the biggest differences between Starship and a conventional mobile robot. Starship’s current commercial pages describe a managed delivery deployment in which the company can map the service area, install charging infrastructure, integrate software, oversee the fleet, maintain the robots and provide ongoing operational support.

The correct commercial question is therefore not simply:

“How much does one Starship robot cost?”

A better question is:

“What will a Starship deployment cost per completed delivery in my operating area?”

That is the metric that matters when comparing autonomous delivery against human couriers, customer collection, vans or another robot platform.

What does Starship say about delivery economics?

In its June 2026 strategy announcement, Starship said its grocery robots can complete deliveries for US$3–4 less per delivery than traditional courier fulfilment.

That is an important claim, but it is a vendor claim—not a universal savings guarantee.

Your result will depend on:

• Delivery density.
• Average distance from the store.
• Robot utilisation throughout the day.
• Order basket size.
• Customer demand at peak and off-peak periods.
• Labour required to stage and load each order.
• Local operating restrictions.
• Weather and route availability.
• Failed and cancelled deliveries.
• The commercial contract negotiated with Starship.

A better way to request a Starship quote

Ask Starship to model at least three cases:

1. Pilot case: one operating zone and a limited number of stores with conservative demand.
2. Target case: expected mature order volume after customer adoption.
3. Scaled case: a multi-store deployment with high fleet utilisation.

For each scenario request:

• Effective cost per completed delivery.
• Expected deliveries per robot per day.
• Expected robot availability.
• Average delivery time.
• Maximum usable delivery radius.
• Minimum contract volume or commitment.
• Integration fees.
• Mapping and launch fees.
• Charging-infrastructure requirements.
• Maintenance and replacement coverage.
• Remote-support coverage.
• Cancellation and failed-delivery treatment.

Do not compare a theoretical robot hardware price with a courier’s all-in delivery cost. Compare completed-delivery economics against completed-delivery economics.

What Is the Starship Delivery Robot?

The Starship Delivery Robot is a compact six-wheeled autonomous vehicle designed to move food, groceries and small goods over the final part of a local delivery journey.

Unlike road-going autonomous vehicles, Starship robots mainly use sidewalks, pavements and pedestrianised areas. They can cross roads and driveways as required but operate at approximately pedestrian speed.

The current robot is referred to by Starship as its e-model.

It combines:

• A six-wheel electric chassis.
• An enclosed insulated cargo compartment.
• 12 cameras including time-of-flight sensing.
• Radar.
• Ultrasonic sensing.
• Onboard perception and navigation.
• Remote-assistance capability.
• Wireless connectivity.
• Customer tracking and phone-based unlocking.
• Wireless charging infrastructure.

What the Starship robot is

• A Level 4 autonomous sidewalk delivery platform.
• A short-radius alternative to some human courier trips.
• A mature commercial system with millions of completed deliveries.
• A managed fleet service for retailers and delivery platforms.
• A local grocery and hot-food delivery solution.
• A small-goods transport option for large industrial sites.
• A customer-facing delivery channel that can integrate with existing apps.

What the Starship robot is not

• It is not an autonomous road car.
• It is not a long-distance delivery vehicle.
• It is not a large grocery van replacement.
• It is not a heavy-payload AMR.
• It is not normally purchased as a general-purpose development robot.
• It is not designed to climb flights of stairs.
• It is not a general indoor service robot.
• It is not capable of delivering through every geography simply because it has autonomous navigation.
• It does not eliminate store picking, packing and loading labour.

If you are still comparing the category, see current delivery robots at Anton Robots.

How Does Starship Robot Delivery Work?

Starship’s service is designed around a relatively simple customer journey, but significant infrastructure sits behind that experience.

1. The customer places an order

The order can be placed through a retailer’s existing application, a delivery platform or a Starship-based application depending on the deployment.

Starship’s APIs can connect the robot delivery layer to an existing ordering system.

2. Store staff prepare the order

The retailer still picks and packs the items.

This is important for labour modelling. Starship automates the physical last-mile trip; it does not automatically pick groceries from store shelves.

3. A robot is loaded

The packed order is placed inside the robot’s cargo compartment at the store or dispatch point.

The compartment is then secured.

4. The robot selects its route

The robot navigates through a pre-mapped operating area using its onboard perception, localisation and planning systems.

It travels mainly on sidewalks and pedestrian routes, crossing roads and driveways when required.

5. Remote assistance is available

Starship says the robots operate more than 99% autonomously.

A human remote assistant can help if the robot reaches an unusual situation it cannot confidently resolve itself. This is different from a person continuously driving the robot.

6. The customer tracks the robot

The customer can see the robot’s location and expected arrival through the app.

7. The customer unlocks the cargo compartment

When the robot arrives, the authorised customer uses their phone to unlock it.

The lid remains secured during the trip.

8. The robot continues operating

After delivery the robot can return for another order or continue through its assigned operating workflow.

A mature fleet therefore depends on more than robot speed. Dispatch logic, charging strategy, loading time and demand balancing all affect how many useful deliveries one robot can complete.

Starship Delivery Robot Specifications

Starship’s current consumer-facing e-model page does not publish every conventional vehicle specification. It focuses on operating capabilities such as autonomy, sensing, cargo volume, battery life and weather operation.

Starship has also published more detailed dimensions and payload figures for its industrial delivery service. Because hardware revisions and configurations can change, those measurements should be treated as published deployment specifications rather than guaranteed values for every Starship robot worldwide.

Why you should confirm the exact robot revision

Starship’s public information has evolved materially over time.

Older technical documents describe substantially shorter battery endurance than the current 18-hour e-model. This illustrates why buyers should not combine specifications from old pilots with a current commercial proposal.

Ask Starship to attach a technical schedule to the contract covering:

• Exact robot revision.
• Maximum cargo weight.
• Internal cargo dimensions.
• Operating speed.
• Expected operating endurance.
• Charging method.
• Weather operating limits.
• Permitted slope and curb geometry.
• Connectivity requirements.
• Required store infrastructure.

Autonomy, Navigation and Remote Assistance

Autonomy is one of Starship’s strongest differentiators.

Starship states that its delivery fleet has operated at Level 4 autonomy since 2018 and that robots are now autonomous for more than 99% of their journeys.

This does not mean humans are irrelevant.

It means the normal delivery is completed autonomously while remote assistants remain available for edge cases.

What the robot handles autonomously

Within an approved and mapped service area, the robot can handle:

• Localisation.
• Route following.
• Pedestrian detection.
• Cyclist and animal detection.
• Obstacle avoidance.
• Stopping when the route is blocked.
• Sidewalk navigation.
• Road and driveway crossings.
• Curb transitions.
• Delivery arrival.

Starship describes the combined sensor coverage around the robot as a “bubble of awareness”.

What remote assistants do

A remote assistant can support the robot when an unexpected situation requires human judgement.

Examples can include:

• An unusual obstruction.
• Temporary construction.
• An ambiguous route.
• A difficult crossing.
• An environmental situation outside the normal operating pattern.

The objective is not to teleoperate every robot for every metre.

The economic value of the platform depends on one human support system being able to supervise a much larger autonomous fleet.

Does the robot need an internet connection to remain safe?

Starship says critical safety functions run locally on the robot and continue functioning even if the internet connection drops.

That is an important architectural distinction.

A communication outage should not remove basic perception and stopping capability.

It does not mean connectivity is irrelevant to the complete delivery service. Tracking, fleet management, customer applications and remote support still depend on communications infrastructure.

Can a Starship robot drive anywhere?

No.

Autonomous capability does not eliminate geographic deployment work.

Starship assesses driveability and maps the sidewalks around participating stores before launch. Routes, crossings, loading points and restricted areas form part of the operational design.

A robot performing extremely well in Helsinki does not automatically prove that an unmapped neighbourhood elsewhere is suitable.

Payload, Cargo Capacity and Delivery Radius

Starship is optimised for small, frequent local deliveries.

The current product page describes capacity as up to approximately three shopping bags.

Starship’s published industrial specification lists:

• Maximum load: 10 kg.
• Cargo dimensions: approximately 400 × 320 × 340 mm.

This is enough for:

• Top-up grocery orders.
• Takeaway meals.
• Drinks.
• Convenience-store orders.
• Small parcels.
• Industrial samples.
• Spare parts.
• Internal mail.

It is not a replacement for:

• A family-sized weekly grocery order.
• Large catering orders.
• Bulk beverages.
• Heavy industrial components.
• Pallets.

How far can a Starship robot deliver?

Starship says the operating distance varies by area but that robots generally work within approximately two miles of stores.

This is a service radius, not simply a battery-range number.

The 18-hour battery means the robot can operate repeatedly through the day. It does not mean an individual customer should be 18 hours away.

Short delivery radius is fundamental to Starship’s economics:

• Trips remain short.
• Robots can return to useful demand quickly.
• Slow pedestrian speeds remain commercially viable.
• A relatively small fleet can cover a dense local catchment.

Order density matters more than theoretical range

Imagine two stores with identical delivery volumes.

One has customers concentrated within one mile.

The other has customers spread across ten miles.

The first is far more naturally suited to sidewalk delivery robotics.

The right deployment question is therefore:

How many suitable orders occur every hour inside a robot-compatible radius?

That number is more useful than asking how far a robot could theoretically travel on one battery.

Hot, cold and frozen goods

Starship describes the cargo compartment as insulated and designed to maintain hot, cold and frozen goods at the required temperature during delivery.

For ordinary grocery and restaurant delivery this is useful.

For pharmaceutical, laboratory or formally temperature-controlled products, buyers should require validated temperature data for the exact product, packaging, journey length and environmental condition rather than assuming an insulated compartment meets regulated cold-chain requirements.

Battery Life, Charging and Fleet Utilisation

The current Starship e-model is advertised for up to 18 hours of operation on a single charge.

That is a major strength for a small delivery robot.

It allows the fleet to operate through long retail trading periods without requiring a battery swap after every few deliveries.

Wireless charging

Starship has developed onsite wireless charging installations so robots can charge without staff manually connecting a cable.

The company describes overnight onsite charging as part of its current operating system.

This matters because charging labour can quietly destroy the economics of a large robot fleet.

A good autonomous delivery system should automate not just driving but also as much of the energy-management workflow as practical.

Published battery figures vary by configuration

Starship’s industrial delivery documentation has also published a 1,260 Wh battery with more than 12 hours of driving time.

The current e-model page advertises 18 hours.

The difference is a reason to confirm the exact current hardware in the deployment contract rather than treating all historical Starship specifications as interchangeable.

What determines real fleet endurance?

Useful daily operation depends on:

• Trip length.
• Number of road crossings.
• Terrain.
• Temperature.
• Snow and surface resistance.
• Robot waiting time.
• Traffic around loading points.
• Battery age.
• Charging availability.
• Fleet-dispatch strategy.

Do not optimise only for battery life

An 18-hour robot that completes very few paid deliveries is not necessarily more valuable than a shorter-endurance robot with better fleet utilisation.

Track:

• Completed deliveries per robot per day.
• Revenue-producing kilometres.
• Idle time.
• Charging time.
• Waiting time at stores.
• Average turnaround between deliveries.

Those metrics determine whether the fleet is productive.

Weather, Curbs and Sidewalk Mobility

Starship has one of the strongest weather track records in the sidewalk-delivery category.

The company operates in climates ranging from Arizona to Finland and describes the current robots as all-weather capable.

Rain

Starship operates commercial services in rainy environments such as the United Kingdom and northern Europe.

The exact contractual weather limits should still be documented for the proposed deployment.

Snow

Starship has substantial real-world experience in Finland.

The current robot uses winter wheels for snowy conditions, and Starship has continued developing snow-aware routing and winter operation through its Finnish deployments.

This is more meaningful than a single controlled snow demonstration because the fleet operates commercially through northern-European winters.

Can Starship robots climb curbs?

Yes.

The current six-wheel chassis uses a bogie system designed to negotiate curb transitions.

This does not mean every curb is traversable.

A high vertical curb, damaged pavement, construction zone or inaccessible crossing can still require another route.

Can Starship robots climb stairs?

No normal Starship deployment should be designed around stair climbing.

The robot is a wheeled sidewalk delivery platform.

If the delivery destination requires movement through staircases, internal building corridors or multiple floors, another robot architecture may be more appropriate.

Sidewalk quality is part of the deployment

Before launch, assess:

• Sidewalk width.
• Surface damage.
• Slopes.
• Curbs.
• Crosswalks.
• Traffic signals.
• Pedestrian congestion.
• Bicycle traffic.
• Construction frequency.
• Snow-clearing practices.
• Loading areas around stores.

Autonomy software cannot compensate economically for a fundamentally unsuitable delivery network.

Safety, Security and Public-Space Operation

A sidewalk delivery robot operates in one of the most difficult robotics environments: uncontrolled public space.

It must interact with:

• Pedestrians.
• Children.
• Wheelchair users.
• Cyclists.
• Dogs.
• Vehicles.
• Construction.
• Street furniture.
• Unpredictable obstacles.

Starship’s commercial scale is therefore relevant. Millions of autonomous deliveries create far more operational evidence than a short pilot.

Obstacle detection

The current robot uses:

• 12 cameras.
• Time-of-flight sensing.
• Radar.
• Ultrasonic sensors.
• Neural-network perception.

The system can identify surrounding objects and either navigate around them or stop.

Road crossings

Starship reports approximately 125,000 road and driveway crossings per day across its global fleet.

Road crossings are one reason route mapping matters so much.

The safest route is not necessarily the shortest geometric line between store and customer.

Visibility

Starship uses:

• Headlights.
• Rear indicators.
• LEDs.
• Reflective elements.
• A high-visibility flag.

These features make the relatively low robot easier for pedestrians and drivers to identify.

Cargo security

The cargo compartment remains locked during the delivery.

The authorised customer opens it through their phone.

Starship also describes:

• Tamper alarms.
• Sirens.
• Location tracking.
• Onboard cameras.

Accessibility cannot be an afterthought

A robot that is safe from a collision perspective can still create a poor public-space experience if it obstructs a narrow footpath, curb ramp or accessible route.

Every deployment should examine:

• Minimum clear footpath width.
• Wheelchair passing space.
• Behaviour at curb ramps.
• Queueing around store loading areas.
• Robot waiting behaviour.
• Behaviour around visually impaired pedestrians.
• How stranded or stopped robots are recovered.

The real safety target is not simply “the robot does not hit people”.

It is safe and predictable coexistence with everyone using the public space.

Cybersecurity and privacy

Starship robots are connected devices operating cameras and fleet software in public environments.

Enterprise deployments should review:

• What sensor data is recorded.
• Retention periods.
• Access control.
• Encryption.
• API authentication.
• Remote-access security.
• Software-update procedures.
• Incident response.
• Customer-account information.
• Local privacy requirements.

Software, API and Retailer Integration

The strongest argument for Starship is arguably not the robot.

It is the infrastructure around it.

A delivery robot that cannot connect into ordering, dispatch, payment, tracking and customer-support systems creates a new operational silo.

Starship is designed to avoid that.

Retailer app integration

Starship can integrate autonomous delivery into a retailer’s existing application.

This allows customers to remain inside the retailer’s normal digital journey rather than having to build a separate relationship with the robot company.

White-label option

If a retailer does not already have a suitable platform, Starship can provide a Starship-based app experience.

APIs

Starship says its APIs can support functions including:

• Creating delivery jobs.
• Loading robots.
• Receiving delivery-status updates.
• Handling cancellations.

Developer documentation and support are part of the integration process.

Mapping

Starship assesses driveability around participating stores and maps the sidewalks used by the robots.

The company says it can map an entire city in approximately one month.

The exact launch schedule will depend on:

• Geographic size.
• Route complexity.
• Permissions.
• Number of stores.
• Integration work.
• Charging installation.

Charging infrastructure

Starship installs the required charging hardware at operating sites.

Space planning should consider:

• Number of robots.
• Parking footprint.
• Electrical capacity.
• Safe movement of store employees.
• Robot access outside store opening hours.

Training

Staff need to understand:

• How to load a robot.
• How to confirm the correct order.
• What to do if loading fails.
• How to report a damaged robot.
• How customer-service issues are escalated.

Fleet management and maintenance

Starship supervises its robots in the field and handles robot maintenance.

This removes a substantial burden from retailers.

The trade-off is that the client is more dependent on Starship’s operational platform than it would be after purchasing and maintaining its own AMRs.

Support

Starship advertises ongoing operational and customer support.

For an enterprise contract, ask for measurable SLAs covering:

• Robot availability.
• Support response.
• Disabled-robot recovery.
• Service outages.
• Integration incidents.
• Replacement hardware.

What Real-World Starship Deployments Show

Starship’s strongest evidence is scale.

The company is no longer trying to prove that a six-wheeled robot can drive down a sidewalk.

The relevant question is whether the delivery model can produce acceptable economics and reliability in a specific commercial market.

What this evidence proves

• Small sidewalk robots can operate commercially at large fleet scale.
• Level 4 delivery does not require one human driver per robot.
• Consumers can adopt autonomous grocery delivery in meaningful volumes.
• Robot delivery can operate through northern-European winters.
• A retailer can integrate the service directly into its own digital channel.
• Starship has progressed well beyond experimental pilot scale.

What it does not prove

• That Starship will be cheaper in every city.
• That every store has enough delivery density.
• That every sidewalk network is suitable.
• That every local authority will permit operation.
• That three shopping bags are sufficient for every grocery customer.
• That the robot can replace all courier deliveries.
• That a successful Finnish deployment will produce identical economics elsewhere.

The 2026 university exit is also evidence

Starship’s decision to wind down U.S. university operations is important buyer information.

The company built much of its early commercial scale on campuses, but in 2026 chose to redeploy more than 1,200 campus robots into grocery and urban hot-food markets.

Starship says the reason is economic opportunity: grocery is a year-round market with stronger scaling potential than seasonal university contracts.

That does not invalidate the campus technology.

It shows something more useful: even a technically successful robotic deployment has to compete for capital and fleet capacity based on commercial returns.

Buyers should apply the same discipline to their own project.

Starship Delivery Economics and ROI

There is no universal Starship ROI.

Robot delivery economics are determined by a network, not by one robot.

The key equation

A useful simplified calculation is:

Total autonomous-delivery cost ÷ successful completed deliveries = effective cost per delivery.

Compare that with:

• Human courier cost per delivery.
• Third-party delivery-platform fees.
• Customer pickup cost and lost conversion.
• Van-based local delivery.

Metrics that determine Starship economics

The strongest Starship site profile

Starship economics are most likely to work when:

• There are many orders inside a small radius.
• The road and sidewalk network is highly connected.
• Average baskets fit within the robot.
• Customers value rapid local delivery.
• Stores can load robots quickly.
• The fleet can remain busy across the day.
• Human courier costs are relatively high.

The weakest profile

Economics deteriorate when:

• Customers are geographically dispersed.
• Orders are infrequent.
• Most baskets are too large.
• The robot frequently waits for store staff.
• Sidewalk routes are indirect.
• Robots spend a high proportion of time empty.
• Many deliveries require a human alternative.

Run a measured pilot

Do not approve a national rollout because a robot demonstration looks reliable.

Pilot a representative area and measure:

• Cost per completed delivery.
• Deliveries per robot.
• Average delivery time.
• Customer rating.
• Intervention rate.
• Cancellation rate.
• Robot availability.
• Order-size rejection rate.
• Store-labour impact.
• Incremental order volume.

Then scale only if those numbers work.

Best Uses for the Starship Delivery Robot

1. Grocery delivery

Best overall use case.

Short-radius grocery delivery is now the clearest strategic focus for Starship.

The platform is particularly well suited to smaller top-up baskets that can fit inside three shopping bags and where customers live within a few kilometres of a store.

The S Group deployment in Finland provides the strongest public evidence for this model.

2. Hot-food delivery

Restaurants and delivery platforms can use robots for short local orders that would otherwise require a cyclist, scooter or car courier.

Pedestrian speed is viable when the delivery radius is small enough.

Starship has integrated with delivery platforms including Uber Eats and Just Eat in European markets.

3. Convenience and rapid local retail

Convenience stores can use robot delivery for:

• Snacks.
• Drinks.
• Household essentials.
• Small forgotten grocery items.

These orders naturally match the robot’s limited cargo volume.

4. Industrial-site logistics

Starship also uses the platform to transport:

• Samples.
• Spare parts.
• Office supplies.
• Internal mail.

This can remove repeated walking trips from employees on large sites.

For heavier industrial logistics, compare purpose-built AMRs with substantially greater payload.

5. Delivery-platform integration

Companies that already aggregate food or grocery orders can add robots as one mode within a mixed fleet.

A good dispatch system can assign:

• Short local trips to robots.
• Longer trips to cyclists or drivers.
• Large orders to cars or vans.

This may be more efficient than attempting to replace every courier trip with the same vehicle type.

6. Customer acquisition and local branding

Starship robots are highly visible in public.

Retailers can apply branded wraps and use robot personalities or arrival music.

This creates a marketing effect that a conventional courier may not provide.

Treat this as additional value rather than the primary financial justification.

When the Starship Delivery Robot Is Not the Right Solution

Starship should not be selected simply because autonomous delivery is attractive.

Reject or deprioritise it when the operating problem points toward another vehicle class.

• Large weekly grocery baskets: three shopping bags are not enough for many family orders.
• Heavy payloads: the published industrial load is 10 kg.
• Long-distance delivery: Starship’s normal operating model is approximately a two-mile local radius.
• High-speed delivery: the robot moves at pedestrian speed.
• Rural delivery: dispersed customers weaken fleet utilisation.
• Door-to-door apartment delivery: the robot is not primarily an elevator-and-corridor indoor delivery system.
• Stairs: a wheeled Starship robot is not a stair-climbing platform.
• Poor pedestrian infrastructure: missing sidewalks or unsafe crossings can make deployment unsuitable.
• Open robotics development: Starship is not positioned like an SDK-first robot chassis for customer experimentation.
• Hardware ownership requirement: buyers wanting to purchase and independently operate their own fleet may prefer another supplier model.
• Low order volume: idle robots can make autonomous delivery uneconomical.
• U.S. university campuses: Starship announced in 2026 that it is winding down this operating segment.

A sidewalk robot is valuable when many small deliveries originate from the same local nodes and travel through a robot-friendly pedestrian network.

If that pattern does not describe the operation, do not force the technology into it.

Starship vs Serve Robotics, Avride, Cartken, Kiwibot and Coco

The correct Starship alternative depends on whether the priority is grocery scale, higher payload, longer range, U.S. urban delivery, campus service or industrial movement.

Which one should you choose?

• Choose Starship when proven fleet scale, grocery delivery and a managed end-to-end operating model matter most.
• Choose Serve Robotics Gen3 when U.S. urban hot-food delivery, greater speed and longer published range are priorities.
• Choose Avride when you need higher published payload, 50 km battery range and IP66 environmental protection.
• Choose Cartken when the task is enterprise or industrial movement across both indoor and outdoor environments.
• Choose Kiwibot when campus deployment or flexible robot rental is central to the project.
• Evaluate Coco 2 when comparing newer city-scale autonomous-delivery platforms.

Use the Anton Robots comparison tool to compare current delivery robots and service platforms.

Is the Starship Delivery Robot Worth It?

Starship is worth shortlisting when the business has a dense volume of small, short-distance deliveries and wants a managed autonomous-delivery system rather than a standalone robot.

Its strongest advantage is operational maturity.

A competitor can publish a larger payload, longer range or higher speed.

Those numbers do not automatically produce a better delivery network.

Starship has already solved many of the less visible problems:

• Mapping.
• Remote assistance.
• Fleet dispatch.
• Customer tracking.
• Store integration.
• Charging.
• Maintenance.
• Weather operation.
• Road crossings.
• Large-scale fleet support.

Where the value comes from

• Reduced dependence on human couriers for short trips.
• High autonomous-operation percentage.
• Long operating endurance.
• Low-energy electric delivery.
• Strong retailer-app integration.
• Real-time customer tracking.
• Outsourced maintenance and fleet supervision.
• Customer novelty and marketing value.
• Commercial evidence at large scale.

Where projects can underestimate cost

• Store staff time for picking and loading.
• Charging-space requirements.
• Integration engineering.
• Operating-area limitations.
• Orders too large for the robot.
• Fallback courier requirements.
• Low demand outside peak periods.
• Local operational or regulatory requirements.

A practical value test

Before entering a deployment, complete this sentence:

Inside a radius of approximately ______ from our store, we currently complete ______ suitable deliveries per day at an average fully loaded last-mile cost of ______ per order.

If you cannot fill in those numbers, you are not yet ready to compare autonomous-delivery economics.

Starship Delivery Robot Deployment Checklist

1. Define the delivery problem. Grocery, restaurant, convenience retail, industrial logistics or another specific workflow.
2. Map current demand. Plot existing orders by customer location and time of day.
3. Measure suitable order percentage. Determine how many orders fit inside the robot’s cargo and weight limits.
4. Define the target radius. Do not assume every customer inside a geometric circle has a viable sidewalk route.
5. Review sidewalks and crossings. Identify route gaps, narrow areas, construction and inaccessible crossings.
6. Review accessibility. Ensure robot routes preserve safe passage for all pedestrians.
7. Model fleet size. Estimate deliveries per robot per hour and demand at peak periods.
8. Review store operations. Measure pick, pack, stage and loading time.
9. Define app integration. Existing retailer app, delivery platform or Starship-based experience.
10. Confirm charging requirements. Power, physical space, robot parking and overnight access.
11. Define fallback delivery. Decide what happens to large orders, out-of-zone customers and service interruptions.
12. Review data and cybersecurity. APIs, customer information, camera data, accounts and remote access.
13. Confirm commercial terms. Fees, minimum volumes, contract length, support and termination rights.
14. Define SLAs. Robot availability, support response, recovery and maintenance.
15. Set pilot KPIs. Cost per delivery, delivery time, utilisation, interventions, failures and customer satisfaction.
16. Run a representative pilot. Include real peak periods, bad weather and normal customer behaviour.
17. Scale only from evidence. Expand after the economics and service metrics meet written thresholds.

Pro tip: ask Starship to provide anonymised operating distributions—not just averages—for delivery time, robot availability, remote interventions and failed deliveries in a market similar to yours. A 22-minute average can hide a very different customer experience depending on the tail of the distribution.

How to Buy or Deploy the Starship Delivery Robot

Starship should not be approached like a conventional robot e-commerce purchase.

There is no standard public “add to cart” price for one current Starship robot.

The commercial path is closer to deploying a logistics service.

1. Define your application

Prepare:

• Country and city.
• Number of stores or sites.
• Current order volume.
• Current delivery radius.
• Average basket dimensions and weight.
• Existing delivery cost.
• Current ordering application.
• Target launch date.

2. Confirm geographic availability

Starship operates only in selected markets and mapped service areas.

Do not assume that a robot can simply be shipped to a new country and activated.

3. Run a driveability assessment

The operating network must be assessed and mapped.

4. Define software integration

Decide whether Starship connects into:

• Your existing retail app.
• A third-party delivery platform.
• A Starship-provided application.

5. Design store infrastructure

Plan:

• Robot loading location.
• Charging.
• Parking.
• Employee access.
• Robot movement around customers and delivery vehicles.

6. Agree commercial and operating KPIs

Your contract or pilot should define measurable performance expectations.

7. Launch a controlled operating zone

Start small enough to isolate problems but large enough to reproduce genuine customer demand.

Review the Starship Delivery Robot at Anton Robots or contact Anton Robots if you want to compare Starship against other autonomous-delivery platforms before approaching suppliers.

If the requirement is still unclear, use the Find My Robot tool before selecting one platform.

What Is New for Starship in 2026?

2026 is one of the most important strategic years in Starship’s history.

Starship is leaving U.S. university campuses

In June 2026, Starship announced that it would wind down its U.S. university-campus operations.

More than 1,200 robots were scheduled to be redeployed toward grocery retailers and hot-food delivery in urban markets across Europe and the United States.

This is a major shift because university campuses were one of the applications most strongly associated with the Starship brand.

Grocery is now the priority

Starship says grocery delivery offers stronger year-round economics and scaling potential.

The company specifically points to Finland as evidence that autonomous grocery delivery can become a normal retail channel rather than a technology demonstration.

More than 10 million deliveries

Starship’s current website lists:

• 10M+ autonomous deliveries.
• 3,000+ robots.
• Operations across eight countries.

By September 2026, Starship’s company updates reported that the fleet had already passed 11 million deliveries, showing how quickly its cumulative volume is increasing.

Grocery operations have expanded beyond pilot scale

Starship’s current grocery-retail page describes operations across more than 300 locations.

The S Group relationship remains the clearest example of grocery scale.

Starship’s 2026 company updates say the S Group network has continued expanding beyond the earlier published case-study figures.

Urban hot-food partnerships are expanding

Starship announced a major partnership with Just Eat in February 2026, beginning with robot deliveries in Sunderland and plans for additional locations.

This follows Starship’s expansion through other food-delivery platforms and demonstrates its move toward becoming an autonomous transport layer inside existing delivery marketplaces.

What this means for buyers

The 2026 Starship proposition is clearer than before:

Starship is becoming a high-scale infrastructure provider for hyper-local grocery and food delivery, rather than trying to serve every possible sidewalk-delivery market equally.

For grocery retailers, that strategic focus is positive.

For U.S. universities looking for a long-term robot-delivery partner, it means Starship is no longer the obvious candidate it was in previous years.

Starship Delivery Robot FAQ

How much does a Starship Delivery Robot cost?

Starship does not publish a standard per-unit retail price for its current delivery robot. Commercial customers generally engage Starship for a managed deployment rather than buying a standalone robot from an online store.

Can you buy a Starship robot?

There is no standard public retail purchase path for the current robot. Starship’s commercial model centres on integrated autonomous-delivery deployments that include mapping, operations, maintenance and software.

How much can a Starship robot carry?

Starship describes the current robot as carrying approximately three shopping bags. Its published industrial-service specification lists a maximum payload of 10 kg.

How big is a Starship Delivery Robot?

Starship’s published industrial configuration measures approximately 697 × 569 × 571 mm without the flag.

How much does a Starship robot weigh?

The published industrial-service specification lists approximately 35 kg.

How fast does a Starship robot go?

Starship describes the robot as travelling at pedestrian speed. A published industrial specification lists a maximum speed of approximately 6 km/h.

How far can a Starship robot deliver?

Starship says robots generally operate within approximately two miles of stores, although the usable area varies by location.

How long does a Starship robot battery last?

The current Starship e-model is advertised for up to 18 hours of operation on a single charge.

How do Starship robots charge?

Starship uses onsite wireless charging installations for its current fleets, allowing robots to charge without staff manually plugging them in.

How long does a Starship robot take to charge?

Starship does not publish one universal current full-charge time on the main e-model page. Confirm charging performance for the exact robot revision and deployment.

Are Starship robots autonomous?

Yes. Starship describes the robots as Level 4 autonomous and says more than 99% of delivery operation is autonomous.

Are Starship robots remotely controlled?

Not continuously. The robots normally navigate autonomously. Human remote assistants are available to help in unusual situations.

Can Starship robots work without internet?

Starship says critical safety functions run onboard even if internet connectivity drops. The wider delivery system still relies on communications for functions such as customer tracking, fleet management and remote assistance.

What sensors do Starship robots use?

The current public specification lists 12 cameras including time-of-flight cameras, radar, ultrasonic sensors and neural-network perception.

Does the Starship robot use LiDAR?

Starship’s current e-model page does not list LiDAR in its main sensor suite. It lists cameras, time-of-flight sensing, radar and ultrasonic sensors.

Can Starship robots detect pedestrians?

Yes. Starship says the robots identify pedestrians, cyclists, animals and other obstacles and either navigate around them or stop.

Do Starship robots drive on roads?

They primarily travel on sidewalks, pavements and pedestrian areas. They cross roads and driveways when required.

Can a Starship robot climb curbs?

Yes. Its six-wheel bogie system is designed to negotiate curb transitions.

Can Starship robots climb stairs?

No normal Starship deployment should rely on stair climbing. It is a wheeled sidewalk robot.

Can Starship robots work in snow?

Yes. Starship operates commercially in Finland and other cold climates and uses winter wheels for snowy conditions.

Can Starship robots work in rain?

Yes. Starship describes its current robots as capable of operating in varied weather conditions including rain.

Can Starship deliver hot food?

Yes. The insulated cargo compartment is designed for hot, cold and frozen goods, and Starship operates restaurant-delivery partnerships.

Can Starship deliver groceries?

Yes. Grocery delivery is now Starship’s primary strategic growth market.

Can Starship deliver directly to an apartment door?

Usually the platform should be considered an outdoor last-mile service. It is not primarily designed to autonomously enter apartment buildings, use elevators and navigate multiple floors.

How does the customer open the robot?

The cargo compartment remains locked during transit and the authorised customer unlocks it using their phone.

Can someone steal the food?

The robot’s lid remains locked during delivery. Starship also uses tamper alarms and robot tracking. No security system can make theft theoretically impossible, but casual access to the cargo is restricted.

How long does a Starship delivery take?

It depends on the market and trip. Starship’s current grocery page publishes an average city delivery time of approximately 22 minutes from store to door.

Does Starship provide an API?

Yes. Starship says its APIs can create delivery jobs, load robots, provide status updates and manage cancellations.

Can Starship integrate with an existing grocery app?

Yes. Retail integration is a core part of the service. Starship can integrate into an existing platform or provide a Starship-based customer app.

Does Starship need to map the area first?

Yes. Starship assesses robot driveability and maps the sidewalk network around participating stores before service begins.

Who maintains Starship robots?

Starship says it handles maintenance as part of its managed service.

Are Starship robots still operating on U.S. university campuses?

Starship announced in June 2026 that it was winding down its U.S. university-campus operations and redeploying more than 1,200 robots toward grocery and urban hot-food delivery.

How many deliveries has Starship completed?

The current Starship website states more than 10 million deliveries. Company updates in September 2026 reported that the fleet had passed 11 million.

How many Starship robots are operating?

Starship’s current public materials describe a fleet of more than 3,000 robots.

What is the best alternative to Starship?

Serve Robotics Gen3 is a strong U.S. urban-delivery alternative; Avride offers greater published payload and range; Cartken is particularly strong for indoor-outdoor enterprise logistics; Kiwibot remains relevant for campus delivery; and Coco 2 is another urban-delivery platform to evaluate.

Is the Starship Delivery Robot worth it?

Yes, when a retailer has high delivery density, small order sizes and a suitable local sidewalk network. It is less attractive for long-distance, large-basket or low-volume delivery operations.

Final Verdict: Should You Deploy the Starship Delivery Robot?

Shortlist Starship if you need high-volume, short-distance autonomous delivery and want a proven managed platform rather than a robot-development project.

Starship’s strongest feature is not one specification.

It is the accumulated system around the hardware.

The company has:

• Thousands of deployed robots.
• More than ten million completed deliveries.
• Level 4 autonomy.
• More than 99% autonomous operation.
• Retailer-app integration.
• Mapped public-sidewalk operations.
• Remote assistance.
• Wireless charging.
• Fleet maintenance.
• Real-world operation through rain and snow.

That is difficult to replicate by simply purchasing an autonomous robot from another manufacturer.

The limitations are equally clear.

Starship carries relatively little cargo, moves at pedestrian speed, operates over a short local radius and depends on suitable pedestrian infrastructure. It also gives the customer less direct hardware and autonomy-stack ownership than an open AMR platform.

The 2026 strategy shift makes the buying decision even clearer.

Starship is concentrating on grocery and urban hot-food delivery. If your operation matches that pattern—many small orders, dense local customers and expensive human last-mile fulfilment—Starship is one of the strongest platforms to evaluate.

If your requirement is heavy payload, long distance, multi-floor indoor delivery or a robot you own and program yourself, choose another platform.

The smartest evaluation path is therefore not:

“Is Starship a good robot?”

It is:

“Can Starship complete enough of our existing local deliveries, reliably and at a lower all-in cost, without damaging the customer experience?”

Build the pilot around that question and the answer becomes measurable.

Ready to compare options? View the Starship Delivery Robot at Anton Robots, browse other delivery robots or request help comparing suppliers and deployment models.