Control · Model Providers
Model Gateway
Auvira sits in front of every downstream model provider and applies one consistent policy framework. Connect providers as governed integrations — the rules that decide what your AI systems can do never change with the vendor behind them.
Organisation Policy
One consistent framework
Policy bundles, data-classification rules, and tool authorisation are defined once and enforced on every provider identically.
Organisation policy remains independent of the underlying model provider.
Providers Active
4
6 integrations
Models Governed
16
across all providers
Policy Framework
1
applied everywhere
OpenAI
ActiveFrontier LLM provider · Vendor API
Allowed Applications
Policy Bundle
Global Baseline v4.2Allowed Data Classification
Region
US-East, EU-WestCapabilities
Tool Access
Gated — approval requiredAnthropic
ActiveFrontier LLM provider · Vendor API
Allowed Applications
Policy Bundle
Global Baseline v4.2 + Legal AddendumAllowed Data Classification
Region
US-East, EU-WestCapabilities
Tool Access
Gated — approval requiredFrontier LLM provider · Vendor API
Allowed Applications
Policy Bundle
Global Baseline v4.2Allowed Data Classification
Region
US-Central, EU-WestCapabilities
Tool Access
Gated — approval requiredAzure AI
ActiveCloud-hosted models · Cloud region
Allowed Applications
Policy Bundle
Global Baseline v4.2 + Data Residency (EU)Allowed Data Classification
Region
EU-West (Ireland)Capabilities
Tool Access
Gated — approval requiredAWS Bedrock
RestrictedCloud model marketplace · Cloud region
Allowed Applications
Policy Bundle
Global Baseline v4.2 + Data Residency (US)Allowed Data Classification
Region
US-East (N. Virginia)Capabilities
Tool Access
Gated — approval requiredSelf-hosted models
Not connectedOn-premise / VPC · Private infrastructure
Allowed Applications
None — awaiting connectionPolicy Bundle
Air-gapped Baseline (draft)Allowed Data Classification
Region
On-premise (self-managed)Capabilities
Tool Access
Tool access denied