Application
Inference & Serving
Portkey Gateway
1Fast 2AI 3gateway 4routing 5many 6LLMs
the six Ws · specification
W1
Who
Maintained by Portkey AI.
W2
What
A fast open-source AI gateway routing requests to many LLMs with load balancing, fallbacks, and caching.
W3
Where
A self-hostable gateway (Node/edge) or Portkey's managed cloud.
W4
When
Reach for it when you need one reliable, provider-agnostic LLM endpoint.
W5
Why
Adds resilience and unified access across models with minimal code changes.
W6
With
Speaks the OpenAI-compatible API; proxies OpenAI, Anthropic, Google, and Azure.
W7
Watch
Maintained by Portkey AI; open-source (MIT) gateway that is self-hostable. Self-hosted routes directly to providers; the paid cloud sees request metadata (open-core).