Sub-processors

Last updated: 17 August 2026.

WebRobot relies on the providers below to deliver the service. The list separates those that process data for every customer from those that come into play only when the customer enables the corresponding feature: if you run no stages backed by language models, no model provider processes your data.

Always involved

ProviderRoleLocationData processed
Hetzner Online GmbHCluster and database hostingGermany / Finland (EU/EEA)All platform data, at rest and in processing
netcup GmbHDedicated compute node hostingGermany / Austria (EU/EEA)Data processed by jobs assigned to those nodes
Stripe Payments Europe, Ltd.Payments and invoicingIreland (EU/EEA)Billing and contact data; no pipeline data

Only when the customer enables the feature

ProviderRoleLocationData processed
Anthropic PBCLanguage models (agents, chat, LLM stages)United StatesContent sent to stages that use a model
OpenAILanguage models, when selectedUnited StatesContent sent to stages that use a model
Together AI · Replicate · RunPodModel training and inferenceUnited StatesTraining datasets and prompts, only for jobs the customer starts
LlamaCloudDocument indexing for retrieval (RAG)United StatesDocuments uploaded to knowledge bases
Hugging FaceDataset and model publishingUnited States / FranceOnly what the customer explicitly chooses to publish
DataImpulseProxy network for data collectionEstonia (EU/EEA)Request destination addresses; not the results

Transfers outside the EEA

Providers based in the United States process data outside the European Economic Area. A customer who must avoid this can leave those features disabled, or run the platform on their own infrastructure, where compute and storage stay inside the perimeter they choose.

Changes

This page is updated before a new sub-processor enters service. To be notified of changes, write to security@webrobot.eu.