Seranai AI Apps · In development
Access your models. Manage usage and cost.
Connect your applications to the AI models you choose, with one place to manage access, routing, usage, and cost.
One connection. More model choice.
Unified model access
Use a common API for supported providers, including OpenAI, Anthropic, Gemini, and cloud-hosted models.
Model aliases
Give applications stable model names while managing the underlying deployments centrally.
Local and hosted models
Connect supported self-hosted endpoints alongside external model services.
More than chat
Access supported embeddings, image, and audio endpoints according to provider capabilities.
Keep requests moving
Routing and load balancing
Distribute requests across configured model deployments using the routing strategy you choose.
Retries and fallbacks
Retry eligible failures and switch to an agreed alternative when a deployment is unavailable.
Rate limits
Set request and token limits to manage demand.
Response caching
Reuse eligible responses where caching fits the application and data policy.
Control access and spending
Application keys
Issue and revoke virtual keys for applications and teams.
Model permissions
Limit which configured models a key can access.
Budgets
Define spending allowances for keys, users, and teams.
Usage and cost tracking
Review token consumption and estimated spend by model and configured account.
Understand the operation
Management dashboard
Manage configured models, keys, and usage from a central interface.
Logging and monitoring
Connect observability tools to investigate requests, errors, and performance.
Configured for your business
Seranai sets up the agreed providers, access rules, routing, budgets, and monitoring, then maintains the gateway as part of your service scope. Logging, retention, and caching are configured around your data-handling requirements. Enterprise features and external services are scoped separately.
Your choice, with clear boundaries
You control your gateway configuration and customer-specific workflows. External model providers retain their own terms and data policies. Model usage charges are separate from gateway service costs unless an allowance is explicitly included. Fallbacks are configured choices, rather than a guarantee of identical answers or lower costs.
Part of Seranai AI Apps
AI Gateway is offered within Seranai AI Apps, separately from Seranai Suite. We are defining the launch scope; supported models, deployment options, and service commitments are confirmed with you.
Tell us how your team uses AI
Share the providers you use, the applications calling them, and where cost, reliability, or access management is becoming difficult.
Questions and answers
What is AI Gateway?
A shared entry point for supported AI models, with routing, access controls, budgets, and usage visibility.
Are model usage costs included?
Model usage is separate from gateway service costs unless an allowance is explicitly included in your agreement.