Which models can run in Europe on Standard Compute?

Claude Opus, Sonnet and Haiku 5.5, GPT models on Azure in Europe, Gemini 3.5 Flash-Lite, and open models including GLM, DeepSeek and MiniMax. With a Europe rule, a model with no approved European endpoint is refused before anything is sent.

Do fallbacks ever leave Europe?

No. With a Europe rule, the classifier, the answer, retries and fallbacks all go only to approved European inference endpoints.

Where are account and usage records stored?

Your account and usage records are stored in Frankfurt, Germany, and our EU gateway runs in Finland. Request records hold metadata only (model, provider, tokens, cost and status), never your prompts or outputs.

Is a Europe rule an extra cost?

No. Region rules and zero data retention are free on every company account.

Who processes our data?

We publish every sub-processor with its purpose, data and location on our sub-processor list, and give you 14 days' notice before adding a new one.

Azure AI Foundry, Bedrock and Vertex AI data residency

Standard Compute keeps your AI requests in Europe across Azure, Amazon Bedrock and Google Vertex AI with one setting. Set a Europe rule for your company, a team or a single key, and every request, including retries and fallbacks, is served only by approved European inference endpoints.

Each cloud handles residency its own way, through deployment types, inference profiles or endpoints. With Standard Compute you set the rule once and it holds for every model and every cloud behind your key.

By Standard Compute · Last updated 9 November 2026

Cloud AI data residency compared

How each cloud keeps inference in a region, from its own documentation, and the Standard Compute equivalent.

Cloud AI data residency compared
	How residency is set	Global option	Regional option
Azure AI Foundry	How residency is set: Deployment type	Global option: Global types may process data in any Azure region	Regional option: Data Zone types (US, EU, APAC) or Standard in one Azure geography
Amazon Bedrock	How residency is set: Inference profile	Global option: Global profiles route to any supported commercial Region	Regional option: Geographic profiles route within the US, EU or APAC
Google Vertex AI	How residency is set: Endpoint	Global option: The global endpoint picks the region and does not meet ML processing requirements	Regional option: Multi-region endpoints (us, eu) or a regional endpoint
Standard Compute	How residency is set: A Europe rule per company, team or key	Global option: Without a rule, routing picks the cheapest provider fast enough to serve each step	Regional option: Approved European endpoints only, including retries and fallbacks

Cloud facts checked on 9 November 2026 against the sources below.

Create a company workspaceSee pricing

Azure AI Foundry data residency

In Azure AI Foundry, the deployment type decides where prompts are processed. Global types may process data in any Azure region, Data Zone types keep processing inside the US, EU or APAC data zone, and Standard deployments keep it inside one Azure geography. Data at rest stays in the resource's geography.

Through Standard Compute, GPT models with a Europe rule are served on Microsoft Azure in Europe, and you never pick a deployment type yourself.

Amazon Bedrock data residency and region availability

On Amazon Bedrock, the inference profile decides where a request runs. A geographic cross-Region profile keeps processing inside the US, EU or APAC; a global profile can use any supported commercial Region and costs about 10% less.

Through Standard Compute, Claude Opus, Sonnet and Haiku 5.5 with a Europe rule are served from Amazon Bedrock in Ireland and Google Vertex AI in Europe.

Vertex AI data residency

On Vertex AI, the endpoint decides. Google advises against the global endpoint when you have ML processing requirements, and offers multi-region endpoints for the US and the EU that keep ML processing inside that boundary.

Through Standard Compute, Claude and Gemini models with a Europe rule are served only from approved European endpoints, and a model with no European endpoint is refused before your prompt is sent.

One Europe rule across all three clouds

Set a Europe rule for your company, a team or a single key, and every request, including retries and fallbacks, is served only by approved European inference endpoints. If no European endpoint can serve it, the request is refused before your prompt is sent.

Our EU gateway terminates TLS and runs in Finland, and your account data is stored in Frankfurt.

Turn on zero data retention for your company, a team or a key, and every request is routed only to inference endpoints that commit to keeping no copy of your data.

Europe here means the EU and EEA, the United Kingdom and Switzerland. Every request record shows the region rule in force and the provider that served it.

OPENAI_BASE_URL=https://eu.stdcmpt.com/v1
OPENAI_API_KEY=YOUR_EUROPE_RULE_KEY

Europe-rule keys call the EU gateway

Questions from IT, security and finance

Which models can run in Europe on Standard Compute?: Claude Opus, Sonnet and Haiku 5.5, GPT models on Azure in Europe, Gemini 3.5 Flash-Lite, and open models including GLM, DeepSeek and MiniMax. With a Europe rule, a model with no approved European endpoint is refused before anything is sent.
Do fallbacks ever leave Europe?: No. With a Europe rule, the classifier, the answer, retries and fallbacks all go only to approved European inference endpoints.
Where are account and usage records stored?: Your account and usage records are stored in Frankfurt, Germany, and our EU gateway runs in Finland. Request records hold metadata only (model, provider, tokens, cost and status), never your prompts or outputs.
Is a Europe rule an extra cost?: No. Region rules and zero data retention are free on every company account.
Who processes our data?: We publish every sub-processor with its purpose, data and location on our sub-processor list, and give you 14 days' notice before adding a new one.