Get started with Hyperlake.
Tell the team what you want to deploy, whose environment it must run in, and which data and policies it needs, and get free expert help with your use case. If you already have a workspace, sign in.
Free expert help
Tell us what you want to deploy
We offer free expert help with your use case, by chat or through the contact form. Tell the Hyperlake team what you want to deploy, whose environment it must run in, and which data and policies it needs.
Who it is for
Enterprise and platform teams
Teams that want to own their AI environment: the data, models, applications and keys stay within infrastructure they control.
AI product companies and service providers
Companies that deploy into many client environments and need repeatable installs, client ownership and one operating approach.
What to bring
What you want to deploy
A private model endpoint, an enterprise agent, a governed data and knowledge foundation, a Physical AI environment. The platform is shaped around the real workload; an example is a private model endpoint with governed access to asset data and a vector store in the customer's environment.
Whose environment it must run in
Yours or a client's. Data, applications, models and keys stay within infrastructure the customer or their client controls, and Hyperlake coordinates the services that run there.
Which data and policies it needs
Your identity provider and the access rules that apply. People and AI agents reach data through one identity path (sign-in through an OAuth/OIDC proxy, scoped identity, policy decisions and audit), not through shared database passwords.
What you get
Less rebuilding
Identity, data services, models, applications, policies and monitoring are packaged together, so each project or client does not start with a new integration.
Less manual work
AI-assisted planning through the same app, CLI and MCP tools to propose, review, deploy, inspect and troubleshoot changes, with human approvals where policy requires them.
Less idle spend
Eligible inference endpoints can scale down when idle, capacity is sized to the workload, and Hyperlake adds no compute markup. Savings come from reuse, reduced integration effort and workload-aware capacity, not a fixed percentage.
How to start
- Read the use cases and the FAQ if you want the detail first.
- Tell the Hyperlake team what you want to deploy, whose environment it must run in, and which data and policies it needs. The help with your use case is free.
- Already have a workspace? Sign in at app.hyperlake.cloud. The sign-in page also lets you start a new workspace with Google or LinkedIn.
Try it first
Free tools that run in your browser
No account needed: compare calling a hosted LLM API with running models yourself in the cost calculator, or explore the open-source Physical AI stack map.
Questions before you start
How is Hyperlake paid for?
Some services, including AI features, agents, paid tools and managed environments, are charged in credits. Managed infrastructure is paid as your plan or order form sets out, and in customer-cloud modes you pay your cloud providers directly. Free credits, when we issue them, expire 30 days after issue. The Terms of Service (section 8) have the details; to discuss costs for your workload, ask the team.
What does the Hyperlake platform include?
Modular capabilities for data and knowledge, models and AI services, and applications and workflows, with shared controls for security and access, observability and resilience, and lifecycle and audit. Teams can also bring their own software.
How do people and AI agents get access to data?
Through one identity path instead of shared, long-lived database passwords: sign-in through an OAuth/OIDC proxy connected to your identity provider, a validated JWT that carries scoped user or workload identity, OPA policy decisions, and enforcement and logging at integrated access points.
Start with a workload. Build the environment around it.
Tell us what you need to deploy, whose environment it must run in, and what it needs to connect to.