Route every AI model through a single API

OpenAI‑compatible endpoint with intelligent routing, automatic failover, and real‑time cost tracking - across OpenAI, Claude, Gemini, and 23+ models.

Evaluate FreeRead Docs
$5 free creditsNo card requiredOpenAI SDK compatible
RoutePlex Gateway

Integration

Drop-in OpenAI SDK. Any model.

Use the OpenAI SDK you already know. Just change the base URL. Access 15 models across 3 providers with built‑in routing, cost tracking, and failover.

  • OpenAI SDK compatible. Change one line of code
  • Also works as plain REST: Python, Node, Go, cURL
  • Per-request cost breakdown in every response
routeplexcoming soonrouteplexcoming soon
app.py
  from openai import OpenAI

  client = OpenAI(
      base_url="https://api.routeplex.com/v1",
      api_key="sk_live_your_key",
  )

  response = client.chat.completions.create(
      model="gpt-4o-mini",  # or "routeplex-ai"
      messages=[{"role": "user", "content": "Hello!"}],
  )

  print(response.choices[0].message.content)

Routing Modes

Three ways to route your requests

From fully automatic AI routing to precise model targeting and fault-tolerant fallback chains.

RoutePlex AI

Recommended

Our router analyses your request (message length, complexity, and your chosen strategy) and selects the optimal model automatically. Balance cost, speed, or quality.

model="routeplex-ai"

Direct Model

Precise

Send requests directly to a specific model when you know exactly what you need. Full control over provider and model selection, zero overhead.

model="gpt-4o"

Fallback Chain

Resilient

Define an ordered list of models. If the first fails, the request silently retries on the next, giving you near-zero downtime across providers.

model="gpt-4o→claude→gemini"

How it Works

From zero to production in four steps

01

Create your API key

Sign up, create an account, and generate your RoutePlex API key from the dashboard in seconds.

02

Send your first request

Point the OpenAI SDK at api.routeplex.com/v1, or POST directly to /api/v1/chat. Works with any language or HTTP client.

03

Choose your routing mode

Let RoutePlex AI pick the best model, target a specific one, or set up a fallback chain for resilience.

04

Monitor & optimise

Track every request, token, and dollar in real time. Set cost caps, rate limits, and budget alerts.

Capabilities

AI gateway features built for production

Intelligent routing, real-time web context, built-in safety, and cost governance. All through one API.

Smart Routing & Failover

Four routing strategies (cost, speed, quality, balanced) with automatic multi-provider retry. If one goes down, requests reroute in ~200ms.

Web-Augmented AI

Prompts are auto-analyzed for search intent and URLs. Real-time web results and page content are fetched and injected into context. Zero config.

Built-in Safety

Three-layer moderation pipeline: pattern detection, AI classification, and URL blocklist. Every request is screened before it reaches any model.

Cost Governance

Real-time cost tracking with micro-cent precision, daily spending caps, budget alerts, and a free estimation endpoint.

OpenAI SDK Compatible

Use the official OpenAI SDK. Just swap the base URL and access all 23 models across OpenAI, Anthropic, and Google through one endpoint.

Real-Time Analytics

Live dashboards for requests, tokens, latency, and error rates. Per-key usage breakdowns and exportable analytics.

Try the AI playground in your browser

Send real requests, compare models side-by-side, and preview costs. No sign-up required.

Open Playground

Trust & Reliability

Reliable AI infrastructure you can depend on

Every layer is designed for uptime, security, and visibility.

99.9%+ Effective Uptime

Multi-provider routing means if one provider fails, your request is silently retried. Your users never see an error.

Automatic Failover

Built-in retry logic with configurable fallback chains. Requests reroute across providers in milliseconds.

Stateless by Design

Prompts and responses are never stored, logged, or written to disk. Data flows through in-memory and is immediately discarded.

OpenAI SDK Compatible

Use the OpenAI SDK with any model. Just change the base URL. Same code, same auth pattern, all 23 models.

Encrypted End-to-End

All traffic TLS 1.3 encrypted. API keys hashed with bcrypt. Fully GDPR-aligned with clear data handling policies.

Granular Rate Limits

Account-level rate limiting, daily cost caps, and configurable token quotas protect your budget.

Pricing

Transparent, usage-based pricing

Start free with $5 in credits. Pay only for the tokens you use. No surprises.

Free Trial

Explore the platform

$0/forever
  • $5 free credits
  • All standard models
  • RoutePlex AI routing
  • 1,000 RPM
  • Basic analytics
  • Premium models
  • Priority support
Evaluate Free
Most Popular

Pay As You Go

Scale with zero commitment

Usage/based
  • Pay only for what you use
  • All standard models
  • Premium models available
  • 10,000 RPM
  • Advanced analytics
  • Cost controls & alerts
  • Email support
Get Started

Enterprise

Large-scale deployments

Custom/pricing
  • Volume discounts
  • All models included
  • Unlimited RPM
  • Custom rate limits
  • SLA guarantee
  • Dedicated support
  • Custom integrations
Contact Sales

Estimate endpoint is always free. View full pricing & model costs →

Frequently asked questions

Quick answers to what developers ask most.

Ready to simplify your AI stack?

$5 in free credits. No card required. Go from zero to production in under five minutes.

Evaluate FreeRead the Docs