o10

AI model and provider integrations

Connect model services to your application with a clear view of credentials, request handling, and operational controls.

Use these guides to understand an integration’s role. Verify endpoint support, account permissions, and current provider terms before sending traffic.

Venues4 integrations
Unified inference gateway + o10

o10 sits above the unified inference gateway, routing to cheapest compliant models while preserving the developer experience of a unified gateway.

OpenRouter + o10

OpenRouter aggregates providers; o10 adds eval-gated routing, shadow mode, and spend enforcement across OpenRouter and other venues.

Amazon Bedrock + o10

o10 routes compliant workloads through Bedrock per-token and committed capacity. Drawing down existing AWS commitments.

Owned / Open-Weight + o10

Self-hosted open-weight models integrate as a venue in o10's routing graph with eval-gated selection alongside cloud APIs.

01Deep dive

Where to start

Begin with a small test request using your own credentials. Check error handling, timeouts, supported features, and usage reporting before expanding traffic.

o10Set the envelope. o10 holds it.

See what you're overpaying.

Paste a week of traffic. Get the number that books the audit.

See what you're overpaying