# Portkey AI Gateway — Route to 250+ LLMs

> Portkey AI Gateway routes to 250+ LLMs with sub-1ms latency, 40+ guardrails, retries, fallbacks, and caching. 11.1K+ stars. Apache 2.0.

## Install

Save in your project root:

## Quick Use

```bash
# Run locally
npx @portkey-ai/gateway
# API at http://localhost:8787/v1
# Console at http://localhost:8787/public/

# Or use Python SDK
pip install portkey-ai
```

---

## Intro

Portkey AI Gateway is a blazing fast gateway with integrated guardrails that routes to 250+ language, vision, audio, and image models across 45+ providers. With 11,100+ GitHub stars and Apache 2.0 license, it delivers sub-1ms latency with a 122KB footprint. Features automatic retries, fallbacks, load balancing, 40+ pre-built guardrails, smart caching, multi-modal support, agent framework integrations, and enterprise compliance (SOC2, HIPAA, GDPR).

**Best for**: Teams needing a unified API gateway across multiple LLM providers
**Works with**: Claude Code, OpenAI Codex, Cursor, Gemini CLI, Windsurf

---

## Key Features

- 250+ models across 45+ providers
- Sub-1ms latency, 122KB footprint
- Retries, fallbacks, load balancing
- 40+ input/output guardrails
- Smart caching (simple + semantic)
- SOC2, HIPAA, GDPR compliance

---

### FAQ

**Q: What is Portkey AI Gateway?**
A: LLM gateway with 11.1K+ stars routing to 250+ models. Sub-1ms latency, guardrails, caching, enterprise compliance. Apache 2.0.

**Q: How do I install it?**
A: npx @portkey-ai/gateway. Or pip install portkey-ai for Python SDK.

---

## Source & Thanks

> [Portkey-AI/gateway](https://github.com/Portkey-AI/gateway) — 11,100+ GitHub stars

---
Source: https://tokrepo.com/en/workflows/585d3a26-0cca-47cb-ac88-2797a853367d
Author: AI Open Source