LibraOS Runtime
No license cost
Developers can build, deploy, and operate AI applications on their own infrastructure without per-agent, per-execution, or per-token Runtime license fees.
Download RuntimeBuild your first appLibraOS Runtime currently has no license cost for development and standard production use, subject to the applicable terms. Plan hosted development, model usage and infrastructure separately.
LibraOS Runtime is currently available at no license cost for development and standard production use, subject to the applicable license terms.
No license cost
Developers can build, deploy, and operate AI applications on their own infrastructure without per-agent, per-execution, or per-token Runtime license fees.
Download RuntimeBuild your first appLibraOS may introduce optional paid enterprise capabilities, support plans, and OEM licensing in the future.
MegaNova Gateway, managed cloud services, and hosted AI development execution are billed separately where applicable.
Understand service costsOur goal is to make LibraOS Runtime accessible to every AI product builder, from independent developers to enterprise software teams.
Choose how your team will develop the application and what help the project needs.
Private beta
Access is enabled per account. Confirm build usage, source export and available deployment options during onboarding. General subscription prices are not published.
Creating a Console account or using the free Runtime does not automatically provide a Cloud workspace.
Request Cloud accessAlready enabled? Open Cloud ↗Scoped engagement
Discuss the client workflow, reusable starting point, acceptance criteria and handover. Assisted pilots start at US$3,000, one-time.
A written SOW defines scope and schedule. This fee covers implementation services; standard production Runtime use currently has no license cost under the applicable terms. Model usage and infrastructure are separate.
Explore the Agency pathDiscuss a client projectExplore the docs, SDK and runnable examples, or develop locally. Runtime development and standard production use currently have no license cost, subject to the applicable terms. Review SDK, starter and third-party component licenses separately.
Start with the developer guideBuilding the application and running it create different costs.
MegaNova Gateway uses separate model-usage billing and account limits. Bring your own provider or compatible local model server if that fits your deployment.
Optional managed inference plans start at $249/month by quote. Confirm models, included allowance and capacity. This is separate from Gateway pay-per-use pricing and does not include unlimited calls.
Compare model connections View Gateway model pricing ↗You operate and pay for self-hosted infrastructure. Managed cloud services and hosted AI development execution are billed separately where applicable. Confirm beta access, available deployment options and charges during onboarding.
Plan your deploymentLibraOS Runtime is currently available at no license cost for development and standard production use, subject to the applicable license terms. Developers can build, deploy, and operate AI applications on their own infrastructure without per-agent, per-execution, or per-token Runtime license fees.
No. MegaNova Gateway, managed cloud services, and hosted AI development execution are billed separately where applicable. You also operate and pay for your own infrastructure and any external model services.
No. Cloud is a private beta. Request access before planning a hosted build; an account, free Runtime use or inference plan does not automatically enable it.
Runtime development and standard production use currently have no license cost, subject to the applicable terms. For each client, agree source access, data boundaries, deployment and support responsibilities. Project services and third-party costs are scoped separately.
You control the application code you write. Cloud export depends on account eligibility. SDK, starter, model and Runtime licenses remain separate, and model compatibility must be tested for your workflow.