Bitrise Remote Dev Environments - Cloud Macs your coding agents can actually build on
by•
Bitrise RDE gives you cloud Macs & Linux machines that start in seconds, the same machines, stacks and caches as your CI.
Give every agent its machine and run them in parallel via MCP, or attach VS Code, Cursor or Claude Code. Archive when you stop, restore later.
Replies
Dial
Giving every agent its own disposable cloud Mac instead of sharing one box is the right call once you're running more than one or two in parallel, that's exactly where local setups start fighting each other over ports and file locks. What I'd want to understand before adopting this is the cost model at real parallelism - if I've got a swarm of 10-20 agents each with their own machine running most of the day, does the CI-cache-sharing keep the bill sane, or does it scale close to linear with instance count?
Bitrise
@galdayan very interesting question :) Let me check with the tech team
Bitrise
@galdayan
RDE pricing is based on the concurrent number of machines. With Basic and Pro pricing, you get up to 5 concurrent VMs. More details here https://bitrise.io/pricing#rde
Bitrise
@galdayan For more concurrency and M4 Pro VMs, we have enterprise pricing. CI cache sharing can help with build speed, especially for the first build in the VM
Bitrise
@galdayan We spent a lot of time talking to users to make the pricing right :) Also, as you are working on your own startup, you are more than welcome to ask for a longer trial or some startup discount 😀
Bitrise
@galdaya your question goes to the core of opinions how people want to consume and not everyone agrees.
This is why we created the following two models:
in self-serve based on concurrency of VMs you max out at a monthly spend and after that the meter stops for the month. As a heavy user you pay a flat fee essentially.
In enterprise pricing there is a lot of flexibility. Can play commit and pay flat, can pay by the minute. Also bursting over your committed capacity is possible, for seasonal peaks.
Dial
@arpadk That's a cleaner split than most usage-based tools manage - capping the self-serve meter instead of just letting the bill run away is the part that would actually make me trust it as a heavy user. The enterprise flexibility (commit-and-pay-flat vs per-minute) makes sense too since a seasonal-peak team has a very different cost shape than a steady-state one.