Show HN: OpenComputer – an open source alternative to Devin
github.comUtpal here, I'm one of the founders of OpenComputer.
OpenComputer is an open source cloud coding agent for teams that can run in your own cloud account (AWS, GCP, Azure).
We run the agent harness and the sandbox on the same durable VM (architecture in the docs [1]). First commit lands in 24s; the alternatives we benchmarked took 53-72s (benchmarks below [2]).
Two other choices:
- One VM per team, scalable horizontally. Sessions are isolated in git worktrees. - Session files are backed up to a repo after every turn, so any session can resume on another VM.
We noticed that past ~640 session-hours a month, one always-on 4 vCPU VM is cheaper than per-session E2B sandboxes, for example. So at scale, this setup would not only be faster than current SOTA coding agents, but also more cost-efficient.
Feel free to try via cloud for free [4] or self host. We would love any and all feedback, and happy to answer any questions! Full demo below [3].
[1]: https://docs.opencomputer.dev/getting-started/architecture [2]: https://opencomputer.dev/blog/cloud-agents-as-fast-as-local/ [3]: https://x.com/IgorZIJ/status/2121309785882591711 [4]: https://app.opencomputer.dev
we've invested heavily into infra one of the non-trivial things we did there is 2 levels of isolation instead of one there are VMs with fast NVMe disks and then inside VMs lighter rootless podman sandboxes - and each organisation gets a fleet of N VMs dedicated, N>=1
this way the VMs serve as a "soft" concurrency limit and all sessions are still isolated, guaranteed to run on your VM, and start fast
At my organization, all of the skilled engineers use Claude Code and the bad ones use Devin cause it’s “easier”.
It shows in the code thats written
I work at Cognition. This is not true. The picker currently has:
- Normal: Balances cost and intelligence with models like Opus 5.5 and GPT-6.2 Sol
- Ultra: Highest intelligence available with Fable 5.2 and GPT-6.2 Astra
- Lite: Fast lightweight models
- Fusion: Multi-model harness that pairs frontier models like Opus 5.5, Fable 5.2, and GPT-6.2 Astra with a cost efficient sidekick for up to 30% cost savings
And then you can select any of the "preview" models like SWE-2, GPT 6.2, Gemini 4 Neon, Fable 5.2, etc.
The built-in modes are basically just "best model at xyz on cost curve" and then we have individual models people can select.
"Premium markup." If you go past the usage included in a paid plan, extra usage is billed at API pricing. In September Devin also got 30–40% cheaper in Fusion and Normal mode and 15–20% cheaper in Ultra while quality held or improved at every mode.
"Cheap, low-quality default model." Devin isn't tied to one model. Our normal router uses Opus 5.5 at the moment (or GPT-6.2 Sol high on GPT plan), and Ultra uses Fable 5.2 (or GPT-6.2 Astra high on GPT plan), only when someone chooses the light mode specifically are they routed to SWE-2 which is actually a great model trained specifically for software engineering that benches really well and developers really love (and SWE-2 inference / usage is currently free).
"Only the bad engineers use it." Our users include engineers from Nvidia, Exa, Modal, OpenRouter, Mercor, Ramp, and hundreds of other companies with a high bar for engineering talent, and the best ones seem to really love and appreciate the quality and shipping velocity they get with Devin.
I encourage anyone reading this to consider asking someone they know who's used Devin within the past few months for their opinion.
If you invested in the company, or are affiliated somehow, consider disclosing..