For most people, I would wait. For developers, I would buy now.
That might sound like a strange conclusion for a robot that just had an unusually strong launch.
MicroDuck launched on August 27, 2026 at an introductory price of $399 before taxes and shipping. Hugging Face said orders surpassed $2.6 million in the first 24 hours, and reporting after the launch said new buyers were facing an estimated four-to-six-month wait. Pollen's own shipping policy is more conservative: it currently lists MicroDuck as a preorder with first deliveries targeted before Christmas 2026, rather than guaranteeing a specific delivery date.
That changes the buying decision.
Not because four to six months is necessarily a problem.
Because four to six months gives the market time to tell us something that nobody could know on launch day:
What is MicroDuck actually like to own?
And I think that information may be more valuable to a normal buyer than getting the robot as early as possible.
The hardware is interesting. That is not my biggest question.
On paper, MicroDuck is unusually capable for its size and price.
It is a 25 cm, 780 g biped with 15 motors, a front camera, an 8×8 ToF LiDAR, two IMUs, microphones, a speaker, NFC antennas, Wi-Fi and Bluetooth, plus an articulated beak that can grasp objects. Pollen says it comes with built-in behaviors and a gamepad so that it is playable out of the box, while developers can use its open-source software stack to create new behaviors and train reinforcement-learning policies in simulation before deploying them to the real robot.
That is a lot of hardware in a $399 package.
But impressive hardware does not automatically produce a product that you will still care about six months later.
This is the part I think is easy to miss when looking at launch videos.
A robot walking around your desk is impressive.
A robot picking something up is impressive.
A robot falling and recovering is impressive.
The harder question is:
What happens after you have already seen it do those things twenty times?
That is where the software ecosystem starts to matter much more than the initial hardware specification.
I think the software ecosystem is the real product
This is probably the strongest reason I would tell an ordinary buyer to wait.
MicroDuck's long-term value depends heavily on what can be built around it.
Pollen is not positioning it as a closed appliance with a fixed feature list. The company explicitly talks about making physical behaviors easier to share, similar to how people build on one another's work with AI models. It describes MicroDuck as a robot that can be played with immediately but also extended through its SDK, simulation environment and reinforcement-learning tools.
That creates two very different products.
The first is:
A small robot that ships with a set of behaviors.
The second is:
A robot whose capabilities keep expanding because other people keep building things for it.
I would be much more excited about the second one.
Imagine a future MicroDuck that can remember your preferences, interact naturally with you, connect to your home, perform small tasks, respond to what is happening around it, or gain new behaviors through community-created software.
At that point, the robot is no longer just a piece of hardware sitting on your desk.
It starts to become a companion.
But there is an uncomfortable truth here:
We do not know yet how large that ecosystem will become.
And that is exactly why waiting can make sense.
Waiting gives someone else the first six months
For a normal consumer, I see the four-to-six-month period differently from a developer.
You are not just waiting for your order.
You are waiting for everyone else to learn first.
Early owners will discover the things launch pages cannot tell you:
- Which behaviors remain fun after repeated use?
- What becomes boring quickly?
- What breaks?
- How much setup is actually involved?
- Which community projects are genuinely useful?
- Which ideas look impressive but are frustrating in practice?
- How quickly does new software appear?
- Does the ecosystem keep growing, or does the excitement fade?
That information has real purchasing value.
A normal consumer does not need to be first.
They need to know what they are buying.
And four to six months gives the community a chance to answer questions that Pollen cannot answer before people have real experience with the robot.
The first community projects are already showing why the ecosystem matters
The interesting thing is that people are already building around MicroDuck before many buyers have received the hardware.
One recent Home Assistant project is a good example.
A developer who had preordered MicroDuck said the robot would not ship until December, so instead of waiting, they started building a voice satellite that would run on the robot. Their project connects Home Assistant with real LLM services through MCP, while using a fake MicroDuck mode for development and validation.
That is much more interesting to me than another demonstration of the duck walking.
Why?
Because the developer is treating the robot as an interface to software rather than as a finished gadget.
The robot becomes:
voice interface + AI + physical body + access to the environment
That is a very different proposition from:
small robot that performs seven cool behaviors.
And it is exactly the kind of ecosystem development I would want to watch before recommending MicroDuck to a non-technical buyer.
If you do not code, the wait could actually work in your favor
There is an important misunderstanding around open-source robots.
A lot of the excitement naturally comes from developers because they are the people who can immediately do something with an SDK, simulator or training environment.
But a non-technical owner does not need to become a robotics developer for the ecosystem to be valuable.
You can benefit from other people's work.
That is the whole point.
If someone builds a better voice interface, you use it.
If someone creates a new behavior, you try it.
If someone connects MicroDuck to a useful home automation system, you benefit without writing the integration yourself.
If someone figures out how to make the robot remember preferences or respond more naturally, everyone gets a better product.
This is where an open software ecosystem can matter to ordinary users.
You do not have to build the ecosystem to benefit from it.
But you do need the ecosystem to exist.
And right now, it is still being built.
This is also why I would buy as a developer
My answer changes completely if I am the person buying MicroDuck to build things.
If I were a developer interested in physical AI, I would have a much stronger reason to preorder.
Pollen is making the SDK, runtime and reinforcement-learning tooling available openly, and the company is explicitly encouraging developers and makers to create and share new behaviors.
For a developer, the four-to-six-month wait is not necessarily dead time.
It can be development time.
You can start with the simulator.
You can understand the software stack.
You can build a behavior.
You can build an integration.
You can experiment with the interfaces.
Then you get the physical robot later and test what you built in the real world.
The incentive is completely different.
A consumer is waiting for the product to become more mature.
A developer can help make the product more mature.
That distinction matters.
There is another reason I would not rush as a normal buyer
The other risk is competition.
Four to six months is long enough for the desktop robot market to change.
A product that looks unusually compelling today might have a more difficult comparison by the time new buyers receive their units.
Maybe a competitor launches something with better personality.
Maybe another robot becomes easier to use.
Maybe a new AI companion product gets much better software.
Maybe MicroDuck itself gets a major ecosystem breakthrough.
We simply do not know.
That uncertainty is not necessarily a reason to avoid MicroDuck.
But it is a very good reason not to pretend the decision is obvious.
The market is moving.
Your preorder is a bet on what the market will look like months from now.
I would be more worried about novelty than hardware
There is one risk I think deserves more attention than the waiting period itself:
What if the novelty disappears?
A robot has a very easy first impression.
It moves.
It reacts.
It looks alive.
It does something unexpected.
You show it to someone.
You record a video.
You try another behavior.
The first few interactions are almost guaranteed to be interesting.
The real test is what happens after that.
A fixed set of behaviors has a finite novelty curve.
An evolving software ecosystem has the possibility of resetting that curve repeatedly.
That is why I keep coming back to software.
If MicroDuck remains mostly a collection of things you can make it do, it may eventually become another gadget that sits on your desk.
If new software continually gives you reasons to interact with it, the same hardware could become much more valuable over time.
The robot does not necessarily need more hardware to become more interesting. It may need better software.
So, should you preorder?
For a normal consumer, my answer is:
I would wait.
Not because I think MicroDuck is a bad product.
Quite the opposite.
I think it is interesting enough that I would want to see what happens next.
I want to see what early owners say.
I want to see which community projects survive beyond the demo stage.
I want to see whether the software ecosystem starts producing genuinely useful and entertaining capabilities.
I want to know whether people still actively use their MicroDucks months after the first wave of excitement.
And by the time that information exists, the production situation should also be clearer.
You may end up buying a better understood product.
If you are a developer, I would buy now.
The calculation is different.
You are not paying $399 just to receive a robot.
You are paying to participate early.
You can start developing before the hardware arrives, contribute to the ecosystem, and potentially create something that makes the robot more valuable for everyone who buys it later.
For that person, being early is part of the value.
If you are a normal person who simply wants a fun robot, I would wait.
Let somebody else discover the rough edges.
Let the community build the first wave of software.
Let Pollen ship more units.
Then decide.
You may find that the ecosystem has become exactly what you hoped for.
Or you may discover that the robot is still mostly a very clever toy.
Either answer is useful.
My take
If I were looking at MicroDuck purely as hardware, I would already find the $399 price interesting.
But I don't think hardware is the real bet here.
The real bet is the ecosystem.
The hardware gets MicroDuck onto your desk.
The software determines what happens after it arrives.
And for an ordinary buyer, I think that makes waiting rational.
For a developer, it makes waiting much less attractive.
So my recommendation is simple:
Want to use MicroDuck? Wait and watch.
Want to build on MicroDuck? Buy early.
The next six months may tell us much more about the product than the launch itself.
Sources
- Pollen Robotics — Microduck product page
- Pollen Robotics — Shipping Policy
- Pollen Robotics — Introducing Microduck
- Reddit — MicroDuck / Home Assistant hardware tester project
- Business Insider — Microduck launch and reported preorder backlog
