01 / 04
One API. Your choice of models.
Connect OpenAI, Anthropic, Gemini, Azure, and compatible endpoints. Keep a familiar API as your model choices evolve.
Enterprise AI control plane
Every AI model.
Every AI model.
One gateway.
One API for your models. One place to govern access, policies, and spend.
01Unified API
02Access control
03Usage & billing
Your AI infrastructure, connectedRequest flow
Your appsWeb, Mobile, API
Your agentsMCP, Workflows
Your teamTools, Playground
GopoGateAI GATEWAYRouteGovernMeasure
◎
OpenAIChat, Embeddings
✳
AnthropicClaude
✦
GoogleGemini
↳
Your endpointAzure, Ollama, BYOK
Your providers. Your keys. One gateway.
◎ OpenAI✳ Anthropic✦ Gemini◈ Azure◉ Ollama+ OpenAI-compatible APIs
Built around your teamLess infrastructure.
Less infrastructure.
More control.
The operating layer between your applications and AI. Keep the flexibility of your providers with consistent control over every request.
02 / 04
Access with clear boundaries.
Give applications scoped virtual keys. Organize teams, set model permissions, and apply budgets and rate limits.
03 / 04
Context, built into the request.
Attach system prompts, content rules, memory, and knowledge to a policy. Reuse it across keys and model deployments.
04 / 04
Know where your spend goes.
See request usage, provider costs, and organization charges. Manage prepaid wallets and inspect an auditable ledger.
A familiar developer experience
Change your endpoint.
Change your endpoint.
Keep building.
Use the SDKs you already know. GopoGate handles provider translation, credentials, and request policies behind one compatible endpoint.
- 01Connect a provider
Add your credentials and register a model.
- 02Create a virtual key
Choose the organization, models, and limits.
- 03Send your first request
Point your client at your gateway URL.
Your model catalog
The right model for the job.
Explore the models configured in this deployment. Prices below are the configured provider rates per million tokens; organization charges can include a markup.
FAQ
A few good questions.
Start with your own providers. Keep control as you grow.
Can I use my existing provider keys?
Yes. Add provider credentials in the console and register the models you want to expose. Clients use GopoGate virtual keys, keeping upstream provider credentials on the server.
Do I need to rewrite my application?
For supported OpenAI-compatible APIs, update the base URL, virtual key, and configured model name. Provider-specific capabilities and parameters can vary, so test your application's requests in the playground.
Can different teams have different limits?
Yes. Organizations, teams, and virtual keys can have their own budgets and allowed models. Keys can also have per-minute rate limits and an expiration date.
Where does the gateway run?
This is a self-hosted Go deployment. The gateway, console, and landing page are served by the same application. Model requests are sent to the upstream providers you configure.
Are all providers and capabilities enabled by default?
No. An administrator configures provider credentials, models, and policies for this deployment. External integrations such as MCP servers and agents require an available endpoint. The console shows the records and usage actually configured here.