New technology rarely wins on power alone. It wins when people can afford to use it and use it without having to start from scratch.
Personal computers reached that point. So did the cloud.
Quantum computing hasn’t quite gotten there yet.
Developers can already rent time on quantum computers through IBM (IBM), Amazon (AMZN) or Microsoft (MSFT). But access is only part of the equation.
Each platform has its own programming tools and billing rules, which can make moving between systems more complicated and costly.
That means the skills and budget you build on one system don’t carry over cleanly to the next.
It matters for your money even if you never write a line of code.
Quantum computing companies topped $1 billion in revenue for the first time in 2025, according to McKinsey.
The consulting firm projects that figure could hit $4.4 billion by 2028. But it can only happen if businesses can actually put these machines to work.
Now a small company in Casper, Wyo., says it can fix the friction.
Postquant Labs’ new virtual machine lets developers write quantum programs once, run them across vendors, and see job costs upfront.
Postquant Labs launches a one-language quantum tool
Postquant Labs launched its Quip Virtual Machine, or QVM, on Tuesday, Oct. 20, according to a company statement.
The QVP allows developers to write an optimization program once and run it on supported quantum and classical computing hardware, without having to rewrite it for each system.
That includes annealers, the type of machine D-Wave Quantum (QBTS) builds, and gate-based systems like those from IBM and IonQ (IONQ).
Classical computers check the answers. Developers can also publish programs and get paid each time someone else runs them, much like an app store.
“We created the QVM because quantum computing is still fractured between competing instruction sets and assemblers, providing few easy onramps to outsiders,” Postquant Labs CEO Colton Dillion said in the statement.
“Our virtual machine has three major goals: to unify and clarify pricing of quantum operations and memory across architectures, to distribute hybrid computation across available machines on our network, and to provide an easier abstraction for new quantum programmers that doesn’t require them to think directly in quantum gates or qubit layouts,” he said.
“This initial release brings together annealers, gate-based systems, and classical computers under one constraint-based programming language, and marks our first steps toward a unified representation for all of computing,” Dillion added.
Every vendor runs its own meter
Here’s the reality: renting quantum time means signing up for one vendor’s meter. Those meters don’t look alike.
More Quantum Computing:
- Bank of America makes fresh 33% call on surging quantum stock
- IBM CEO sends blunt message on quantum computing
- Nvidia once rattled IonQ stock. Now it plans to install IonQ tech
IBM’s prices apply only to IBM hardware. Postquant Labs claims Microsoft’s Azure Quantum typically requires about $1,000 in spending before a first circuit can run.
Programming is fragmented, too. Each vendor’s software stack works differently, so code written for one machine usually has to be reworked for another.
Here’s what IBM and Amazon actually charge
I ran the published list prices to see what a first project actually costs. The math adds up fast.
Here’s how the numbers stack up, according to IBM’s pricing page and Amazon’s Braket pricing page:
Quantum computing prices at a glance
- IBM pay-as-you-go: $96 a minute, or $5,760 an hour
- IBM Flex plan: $72 a minute, with a 400-minute minimum of $28,800
- IBM Premium plan: $48 a minute, with a 5,200-minute minimum of $249,600 a year
- Amazon Braket: $0.30 per task, plus a per-shot fee that varies by machine
- Amazon’s own 10,000-shot example: $9.30
None of those prices is outrageous for a corporate research lab. The catch is that each one covers a single platform, so comparing them takes real work.
Each job gets a price ceiling upfront
Postquant Labs’ answer is metered pricing with a ceiling. Each job’s cost depends on how many qubits it uses, how long it runs and how much entanglement it needs.
Every job has a base price plus an optional priority fee. What you see is the maximum cost before anything runs.
Postquant Labs Chief Technology Officer Richard Carback calls those metered units “gas.” He argues that predictable pricing is what turns quantum into a business tool.
“Using a pay-as-you-go ‘gas-based’ cost model is the missing piece that demonstrates quantum computing is commercially viable and practical,” Carback said.
“When you can predict exactly what a computation will cost before you run it, quantum tests that began as a research experiment become a real tool for engineering because you can quantify exactly how much value it can bring to your products. This is the shift we will be delivering,” he added.
There’s a catch. The test version comes with free compute credits, but the company hasn’t published what paid jobs will cost.
The 1,930-race demo comes with caveats
Portfolio optimization is the math behind picking an investment mix that balances return and risk. It’s one of the jobs quantum computers are supposed to do well.
At the Quantum.Tech World event, the QVM ran 1,930 races between a quantum solver and a classical solver, according to Postquant Labs’ demo write-up.
About 100 visitors created the problems. Results within 0.5% of each other were counted as ties.
The company claims its quantum solver matched or beat the classical one in every race. When the gap topped 0.5%, quantum built the better portfolio by a median of 2.56%.
In my analysis, those numbers need more context. The company ran the demo itself, and the release doesn’t name the classical solver or explain how “better” was scored.
Speed wasn’t the edge, either. Both solvers typically answered in 0.1 to 0.3 seconds.
Universities kick the tires on the QVM
Researchers at Monash University, the University of Notre Dame and Purdue University are assessing the QVM. Their pitch is about saving time, not raw power.
“Oftentimes, when we put a student on a project, it takes them the whole semester just to learn how to program using the latest quantum paradigms,” Monash University professor Terrence Mak said.
“As we evaluate Quip’s virtual machine, we believe it can speed up student training and give them more time to do real research instead of spending cycles learning a specific programming language or how to program quantum circuits,” Mak added.
“Working with Quip’s QVM can help researchers and students spend less time learning specific quantum programming tools and more time applying quantum methods to practical research problems, including our work on the Lagrangian operator,” University of Notre Dame doctoral candidate Shravan Kumar Sharma said.
SiC Systems wants faster answers
SiC Systems is exploring whether the QVM can handle industrial problems. These include artificial intelligence-driven manufacturing and the control of chemical and biological processes.
“At SiC Systems, we face a variety of combinatorial optimization problems, and time is of the essence. We want to explore as many solutions as possible in as short a time as possible,” SiC Systems Chief Science Officer Soheil Mansouri said.
“This is particularly important in our physics-informed agent architecture to design, control and operate complex chemical and biological manufacturing processes. Quip Network’s QVM can help us meet that requirement and better serve our customers,” he added.
Every job also passes three layers of checks. Quantum results are compared against classical simulations, and classical providers benchmark them independently.
A tamper-evident record then ties each result to the machine that produced it.
Should IonQ and D-Wave investors care?
If you own quantum stocks, the hardware gets the headlines. Software that cuts the semester students spend learning to program could matter just as much.
Easier tools widen the pool of customers who can use the machines. That’s the bull case for IonQ and D-Wave Quantum.
The realistic case is that both companies are still posting operating losses, according to The Motley Fool. Quantum stocks also routinely swing by double digits.
Postquant Labs’ launch is a test release from a small company, and its demo numbers are self-reported. Treat it as a sign of where the industry is heading, not a reason to chase a stock.
The machines keep getting better. The bigger question is whether anyone besides a physicist can afford to use them.
Related: Cathie Wood buys $11.8 million of tumbling tech stock
