Pricing
Pick a delivery mode. With SaaS we host and run everything on our cloud: you pay a monthly canone plus metered, prepaid credits for the compute and AI you actually use — pay-as-you-go, nothing to operate on your side. With BYOC the machines are yours — elastic VMs in your cloud account, or dedicated servers you already own — and you pay a licence per node; we never resell compute, so there is no mark-up on it. Use the toggle below to compare.
Either way it is the same stack and the same agents; what changes is who runs the machines and how you pay for them. The AI usage of your agents is yours too: your model key, or ours within the limits shown at sign-up. On SaaS, the metered usage (compute, storage, proxy and AI) is billed from the underlying cost with a default +30% mark-up — the same unit prices for everyone. If you want to spend less, BYOC runs the compute on your own cloud with no mark-up — the cheapest way to run at scale. (The base Starterplan is SaaS-only; Growth and Pro work both ways. Managed Kubernetes iscoming soon; BYOC Hard is on demand for enterprises and institutions, after the necessary security checks.)
Indicative, from the same rates as the table above. The real bill is metered from actual usage.
Prices exclude VAT: individuals pay the VAT of their country on top, shown at checkout; companies with a VAT ID are invoiced under reverse charge where it applies. Access is by invitation: register with your invite on webrobot.eu or request one.
SaaS or BYOC — how do they bill?
On SaaS (managed) we host and run everything on our cloud: you pay the monthly canone plus metered prepaid credits for the compute, AI and bandwidth a run actually uses — pay-as-you-go, the same unit prices for everyone (switch the toggle above to “SaaS” for an indicative estimate). On BYOC there are no usage credits: the machines are yours and you pay a licence per node; we never resell compute.
What is Managed Kubernetes — and is it SaaS or BYOC?
A dedicated cluster on managed Kubernetes (GKE, EKS, AKS or OVH managed) — no kubeadm, the managed service runs the control plane. It comes both ways, at €199 per node per month: as SaaS we provision and operate the cluster for you, and compute & AI are billed on top as credits; asBYOC it runs on your own cloud account and you pay your provider for the nodes. Either way you get the whole stack — acquisition, ETL, agents, trading — with per-tenant isolation and encrypted runtimes. It is not launched yet: join the waiting list.
What counts as a node — and what does not?
A node is a machine that stays up as part of a dedicated cluster (Managed Kubernetes, BYOC Hard): in BYOC Hard the control plane counts as one, which is why the minimum is three. The ephemeral VMs we create for a single run — the browsers and the Spark executors — are not nodes and are not licensed: they exist while the job runs, and in BYOC you pay your provider for those minutes. Growth and Pro have no nodes at all: only ephemeral VMs, up to 10 and 25.
Why do BYOC Hard clusters start at 3 nodes?
Three nodes give a highly available control plane and replicated storage: with fewer, one failure takes the cluster with it. BYOC Hard is quoted on demand, for enterprises and institutions.
What is in the monthly fee, exactly?
Orchestration: the designer, the scheduler, the engines, the agent runtime, the API and the supervision of every run. Not in it: the machines that execute the work, and the model usage of your agents. On SaaS we run those on our cloud and bill them as metered credits (our cost +30%); on BYOC (and Pro in BYOC mode) the elastic VMs run on your own cloud account and cost what your provider charges, with no mark-up from us.
Do you charge for LLM usage?
No plan includes it. You connect your own model key — or your own hosted models — and pay the provider directly; on Starter, Growth and Pro you can use ours with the limits described at sign-up.
Starter, Growth or Pro?
Starter orchestrates up to 3 machines on the lightweight engine for web pages, and its AI agent answers in chat — on your site or on Telegram. Growth adds the two things that change what you can build: distributed Apache Spark — on any source, databases, files, APIs, documents, IoT and web pages, for volumes a single machine cannot hold — and agents that work in the background on a schedule or on an event instead of waiting to be asked — with up to 10 ephemeral VMs. Pro goes to 25 VMs and adds teams of agents that work together. In every plan the VMs — Spark executors or browsers — are created when a job starts and destroyed when it ends.
Why can I not buy Managed Kubernetes or BYOC Hard from the console?
They are dedicated clusters — a control plane and nodes that are yours and stay up — and we open them one customer at a time, after the security checks a dedicated cluster needs. Managed Kubernetes is coming soon; BYOC Hard is for enterprises and institutions, on demand: talk to us. Starter, Growth and Pro, on ephemeral VMs, are available today.
And if I need you to build it with me?
System integration and vertical development are charged at €100 per hour, on request: connecting the platform to your own systems, porting an existing scraping estate, building a vertical on top of it, or sitting with your team while they learn the designer. It is separate from the plans above and never required to use them — the platform is meant to be driven without us. Ask for an estimate before we start; an open-ended hourly engagement is nobody's interest.
One thing about the rate that is worth knowing: we build with AI in the loop — Claude Code and OpenAI Codex, the same agentic way of working this platform exists to support — so an hour tends to cover more ground than a traditional development hour. That is why the number is what it is. You can judge the method by the product: this platform was built that way.
Are there different terms for the group's startups?
Yes. Startups in the PioneersGalaxy studio get reduced rates, and part of the work can be done for equity rather than for cash — the platform exists in part to make their products possible, so charging them a market rate for building on it would be beside the point. If you are founding something and the fit is real, say so when you write.
Can I sell my own pipelines or agents?
Yes — the developer marketplace, now in construction, will pay a revenue share on every install or run. Learn more.
Air-gapped installs, multi-region, partner white-label or volume licences. Integration and vertical development at €100/hour when you want us building alongside you.
