OSS TanbouOn this page
Overview
Portkey AI Gateway provides an OpenAI-compatible entry point that can route requests to multiple LLM and multimodal providers. It centralizes retries, fallbacks, load balancing, conditional routing, and guardrails and can be deployed with Node.js, Docker, Cloudflare Workers, and other environments.
Features and best fit
Based on official documentation; not hands-on tested · Content checked:
Key features
Unify multiple model providers behind one API
Applications can use an OpenAI-compatible request shape while the gateway connects to providers such as OpenAI, Anthropic, Gemini, and others, reducing provider-specific routing logic in application code.
Sources: [1]
Add retries, fallbacks, and load balancing at the gateway layer
Routing configurations can define automatic retries, failover to other providers or models, weighted load balancing, and request timeouts.
Sources: [1]
Centralize guardrails and operational visibility
Guardrails, logging, and usage analytics can sit at the gateway layer, creating a shared control point across multiple AI applications.
Sources: [1]
Best fit
Fits multi-model applications that want provider differences and failover in a common layer
It is relevant when model switching, fallback behavior, and policy enforcement should not be reimplemented independently in every application.
Sources: [1]
Before adoption
Design provider credential handling before exposing the gateway
Requests require provider credentials. Plan secret storage, access boundaries, logging, and what data is forwarded to external model providers before production use.
Do not treat the 2.0 pre-release as the stable 1.x release line
The latest stable GitHub release is v1.15.2, while the README advertises Gateway 2.0 as a pre-release. Pin the package, branch, or image used for production so the rollout track is explicit.
Review the boundary between open-source, hosted, and enterprise capabilities
The README also describes hosted and enterprise functionality. Confirm that required governance, observability, and security features are available in the deployment model being evaluated.
Sources: [1]
Official sources
- [1]Portkey AI Gateway README(2026-10-03)
- [2]Portkey Gateway deployment guide(2026-10-03)
- [3]Portkey AI Gateway v1.15.2 release(2026-10-03)
- [4]Portkey AI Gateway MIT license(2026-10-03)
Supplemental curator note
Portkey AI Gateway is useful when multi-model applications need provider routing, failover, and policy controls in one layer. The README currently presents Gateway 2.0 as a pre-release, so define clearly whether production workloads remain on the stable 1.x line or move to the 2.0 track.
Try it in 3 steps
- 1
Start the stable gateway
Start stable v1.15.2 in a Node.js/npm environment. The API listens at http://localhost:8787/v1.
npx @portkey-ai/gateway@1.15.2 - 2
Open the Gateway Console
Open the local console and confirm that the gateway is running.
Open http://localhost:8787/public/ - 3
Send one request with your provider key
Provide the credential for the selected provider through an environment variable and verify a response through the gateway.
curl http://localhost:8787/v1/chat/completions -H 'Content-Type: application/json' -H 'x-portkey-provider: openai' -H "Authorization: Bearer $OPENAI_API_KEY" -d '{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Say hello"}]}'
Growth
Growth trends · Last 30 days
13,123 Stars
Trend data is still being collected.
Development activity
Last 90 days · weekly
- Commits (last 30 days)
- 0
- Open PRs
- 160
Development activity is still being collected.
Built with
Categories and tags
GitHub data
GitHub dataView detailed GitHub data
GitHub Topics
- gateway
- generative-ai
- llmops
- llms
- ai-gateway
- langchain
- openai
- hacktoberfest
- llm
- llm-gateway
- mcp
- mcp-client
- Stars
- 13,123
- Forks
- 1,327
- Watchers
- 63
- Open issues
- 124
- Contributors
- 102
- Owner type
- Organization
- Primary language
- TypeScript
- License
- MIT
- Repository last updated
- May 25, 2026
Related information
Write a related articleShare a guide or use case for this OSS in Markdown. Articles are published after administrator approval.
Explore next
- Dify157,734 Stars
5 shared tag(s) · 1 shared category(s) · Same language
design and publish AI workflows, retrieval, and agents on one platform
TypeScript - LangWatch4,911 Stars
4 shared tag(s) · 2 shared category(s) · Same language
run LLM tracing, evaluations, and agent testing in one production platform
TypeScript - LibreChat45,228 Stars
4 shared tag(s) · 1 shared category(s) · Same language
operate multiple AI providers and agent features through one self-hosted chat
TypeScript - Kong Gateway44,238 Stars
4 shared tag(s) · 1 shared category(s)
Centralize API routing, authentication, and traffic control at a shared gateway
Lua - Keep12,374 Stars
4 shared tag(s) · 1 shared category(s)
centralize monitoring alerts and connect correlation, enrichment, and response workflows
Python - LLM Gateway1,671 Stars
3 shared tag(s) · 3 shared category(s) · Same language
route multiple LLM providers through one API while centralizing usage and cost visibility
TypeScript
Report incorrect information
Tell us if any listing information is incorrect or outdated.