box
Simple computers for agent w/ full VMs
Start new thread

box - Simple computers for agent w/ full VMs

by•
Box offers the simplest, cheapest cloud computers for agents, thought for builders of agentic platforms & software factories. Run 'box new' in your terminal, in 2s get a beefy ubuntu VM, with admin rights, a desktop, ssh access. At $0.036/hr it is 10x less expensive than the likes of E2B, Daytona, Modal, so you can run more agents, or run them longer. Run up to 1000 concurrent boxes fully self-served, or ask us for more, with same-day support from the founders.

Add a comment

Replies

Best
Maker
📌

Hello Product Hunt,

2 years ago, I worked at the European Space Agency, scaling their black holes & binary stars research efforts.

At that time, GPT 4o already helped us find many optimizations that made the code more scalable. But to test them, our agents needed a cluster of cloud compute, which we had to wait 2 months to get. AI may work fast but compute provisioning is slowing it down.

After leaving, I spent 10 months building cloud coding agents, which needed cloud computers too, so I tried many self-serve options for "computers for agents" and "AI sandboxes".

But there was no fit, they were optimized for bursty, short-lived scenarios on containers rather than full computers where anything just works...

... all while being super expensive, high multiples of what the same compute is worth in the average datacenter in Europe.

So 2 months ago, we built and launched box on X, our on-demand computers for agents, super cost-efficient and powerful.

Now our computers for agents have already run for 10 years in just 2 months, growing steadily every week.

Today, they already power many agentic platforms, software factories, cybersecurity audits platforms, with real customers in production.

Try box and let us know how it goes :)

 congrats on the launch!

Maker

 thank you Marek!

  Congrats on the launch. You guys built a great product. Happy user here!!

Maker

 thanks Bhanu!

The months to provision a cluster detail is the real story here that's the actual bottleneck most "AI moves fast" narratives skip over. Compute governance and procurement cycles don't care how fast your model iterates. Curious how you're hitting $0.036/hr while staying full VM rather than container is that just leaner infra overhead, or are you oversubscribing hardware more aggressively than E2B/Modal do? Also: what happens if an agent inside the VM does something destructive is there a snapshot/rollback layer, or is isolation the whole safety model?
Maker

 we hit such low price because we designed the whole system for cost efficiency, using no dependencies and minimizing the operational costs. we have our own snapshotting system (beautiful tech built in-house), allowing you to easily stop/resume/fork your machine.

Maker

 Compute in European datacenters is just that cheap, its just hard to build around, we still make v good margins.

Maker

Hey guys!

im Kirill, the CTO of box.

Today, we're seeing 7 consecutive weeks of big usage growth, and compared to six months ago, our MRR is 130× higher. It's been an incredible ride, but we're still early, and we're shipping every single day.

I'd genuinely love your feedback: good or bad. If you're building coding agents, personal assistants, or any agentic product, you can find box super useful!

Maker

 The goat

Full VM + snapshot/fork is the right shape for agents that need real state, not just a clean container every run.

The self-serve scale is the part I’d want to stress test. If someone spins up hundreds of boxes for long-running coding agents, do fork/resume times stay predictable, or does it vary a lot once the fleet is hot?

Maker

 no it's stable, even at the thousands box per user scale

@anic_dev nice. That predictability at thousands-per-user scale is the part that makes this much more interesting than a cheap sandbox. If fork/resume stays stable there, long-running coding agents become a lot easier to design around.

Good luck with the launch guys, amazing product!

Maker

 thank you Nikita!!

Very cool product. Having deployed at scale, most agent sandbox pain is re-provisioning, not compute. Curious how fast fork actually is under load. The persistent-VM + fork-from-snapshot combo is the part that actually changes how you build agent loops.

Maker

 So many hard problems to solve in that space for so many different usecase.

We catter best to long running agent whose data is precious and never to be lost

I'm a box user, been using it for (I think) a few weeks now. It's insanely fast, cheap (actually ridiculous), and good. Try them out! And congrats on the launch guys

Maker

 Thank you so much Kartik! It's a pleasure to help you

too simple to use and spin up - been working on integrating this in my platform.

Maker

 Hope it's going well! Lmk if I can help in any way :)