🔥 Introducing the Helicone AI Gateway, now on the cloud with passthrough billing. Access 100+ models with 1 API and 0% markup.
Stop juggling API keys and invoices. Fund once, use everywhere. Full observability included at exact provider prices.
Â
One integration
Access OpenAI, Anthropic, Google, Meta, and dozens more through a single endpoint. Automatic failover, load balancing, and seamless provider switching.
View documentationReliability
Claude down on Anthropic? We route to Bedrock or GCP. Same model, different provider.
Learn about provider routingIncluded free
Complete Helicone platform included. Analytics, debugging, monitoring, caching, and more.
Explore platform featuresAlways the best price
Automatically routes to the cheapest provider first. Instant failover on rate limits, timeouts, and errors. Zero-config reliability across 100+ models.
Gateway technical docs01
One-time or recurring top-ups. Your balance works across OpenAI, Anthropic, Google, and 100+ other models. No expiration dates, no provider lock-in.
02
One endpoint for all providers. Automatic failover, load balancing, and response caching. Switch models without changing code.
import { OpenAI } from "openai";
const client = new OpenAI({
baseURL: "https://ai-gateway.helicone.ai",
apiKey: process.env.HELICONE_API_KEY,
});
// Works with any model from any provider
const response = await client.chat.completions.create({
model: "o3", // or claude-opus-4, gemini-2.5-pro, grok-4, llama-3.3-70b...
messages: [{ role: "user", content: "Hello!" }]
});03
Monitor usage, debug errors, analyze performance. Set spending limits, share credits with your team. Everything you need to scale — at exact provider prices. Only standard payment processing fees apply.
One integration for 100+ models. Full observability included at exact provider prices.
Be the first to know when Credits launches.