osFoundry / Platform
Pure usage billing
No seats, no plans, no commitments. Just the resources you use, priced clearly.
What it is
Pure usage billing is osFoundry's pricing model: you pay for the resources you actually use, and nothing else. There are no per-seat fees, no monthly plan tiers, and no annual commitments to sign. Add your whole team without watching a headcount meter, and stop paying for capacity you aren't touching. Everything is priced clearly against what ran, so the bill reflects real work rather than a subscription you have to grow into.
Because osFoundry is bring-your-own-key, most of what you'd normally pay a reseller for goes straight to the source. You connect your own API keys for the models you run — local, cloud, or self-hosted — and model usage is billed by those providers at their own rates, with no osFoundry markup layered on top. That keeps the largest line item, LLM inference, entirely under your control and visible in your provider's own dashboard. osFoundry's own charges are limited to the platform resources it actually provides, not a percentage skim on every token you send.
This matters because it removes the two things that make AI tooling expensive and hard to reason about: seat licenses that punish you for having a team, and reseller margins hidden inside model calls. With a local-first workspace, a lot of work can stay on your device and never incur a cloud charge at all. What's left is a transparent, usage-shaped bill: the resources consumed, priced clearly, with no plan to outgrow and nothing to cancel.
Key capabilities
- No per-seat fees — add your whole team without the bill scaling by headcount
- No monthly plans or annual commitments; pay only for the resources you actually use
- Bring your own API keys so model usage is billed by providers directly, with no reseller markup
- Run local, cloud, or self-hosted models — local-first work can stay on your device and off the bill
- Clear, usage-shaped pricing against real resource consumption instead of a subscription tier
Use cases
- Small team lead: onboards five teammates onto one workspace without buying five seats, and pays only for the model calls and resources they actually run
- Cost-conscious builder: routes most inference through their own API keys so LLM usage bills at provider rates with no osFoundry markup on top
- Privacy-focused analyst: keeps sensitive work local-first on their own device, incurring platform charges only for the cloud resources they genuinely use
Frequently asked questions
Does osFoundry charge a per-seat or subscription fee?
No. There are no per-seat fees, no monthly plan tiers, and no annual commitments. You pay for the resources you actually use, priced clearly, so adding people to a workspace doesn't change the pricing model.
Who bills me for the AI models I run?
osFoundry is bring-your-own-key, so you connect your own API keys and model usage is billed directly by those providers at their rates. osFoundry does not add a reseller markup on top of your model calls.
Can I keep costs down by running things locally?
Yes. osFoundry is local-first, so work that runs on your device can stay there and avoid cloud charges entirely. You're billed for the platform resources you actually consume, not for a fixed plan you have to fill up.