UNIFIED AI ACCESS

Flagship models.
20–70% cheaper, every call.

OpenAI-style chat completions and Anthropic-style messages, one endpoint for Claude and GPT. Swap your base URL, keep your SDK, and pay 20–70% less than going direct — no minimums, no rate-limit games.

Get your API keySee how it works
Works withClaude Code·Codex·opencode
// same SDK — just point it at Inferray
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.inferray.com/v1",
  apiKey: "inferray_...",
});

const resp = await client.chat.completions.create({
  model: "gpt-5.2",
  messages: [{ role: "user", content: "ship it" }],
});
OpenAI & Anthropic compatible·Streaming built in·Zero prompt retention·Prepaid — no subscriptions·No monthly minimum
THE MATH

Same tokens.
A 20–70% smaller bill.

We buy inference at volume and pass the discount straight through. You get the exact same models — just billed lighter.

20–70%
off list price
on every model, always
Monthly inference spend$2,000
You save / month
$400–$1,400
You save / year
$4,800–$16,800

Drag to estimate. Applied automatically to your account.

ONE KEY, EVERY FLAGSHIP

Every model, discounted

Anthropic, OpenAI, and Google's newest, side by side, at one endpoint. Prices per 1M tokens (input / output).

MODELPROVIDERDIRECTINFERRAY
Claude Fable 5claude-fable-5Anthropic$10 / $50$7.50 / $37.50Claude Opus 4.8claude-opus-4-8Anthropic$5 / $25$3.75 / $18.75Claude Sonnet 5claude-sonnet-5Anthropic$2 / $10$1.50 / $7.50Claude Haiku 4.5claude-haiku-4-5Anthropic$1 / $5$0.75 / $3.75GPT-5.5gpt-5.5OpenAI$5 / $30$4.00 / $24.00GPT-5.4gpt-5.4OpenAI$2.50 / $15$2.00 / $12.00GPT-5.2gpt-5.2OpenAI$1.75 / $14$1.40 / $11.20Gemini 2.5 Progemini-2.5-proGoogle$1.25 / $10$1.06 / $8.50Save 20–70% on every one of them — explore the full catalogue →
THREE LINES, NOT THREE DAYS

Drop in without rewriting anything

If your code already talks to OpenAI or Anthropic, you're basically done.

01

Swap the base URL

Point your existing client at api.inferray.com and drop in your Inferray key.

02

Keep your SDK & agents

Point Claude Code, Codex, or opencode at Inferray — same tools, same commands, lower bill.

03

Ship & watch the logs

Every request lands in your dashboard with latency, tokens, and cost broken out live.

Streaming built in

Server-sent events pass straight through, token by token, with usage still tracked.

Usage logs & insights

Per-key, per-model spend, tokens, and latency for every request in your dashboard.

Zero data retention

Prompts and completions are never stored or trained on — only usage counts.

PAY HOWEVER YOU WANT

Billing that stays out of your way

Credit card

Add a card, buy credits, done. Itemized invoices and receipts for every top-up.

VISAMCAMEX
RECOMMENDED

Auto top-up

Set a floor and a refill amount. Dip below the threshold and we top you up — no 3am outages.

when balance < $50 → add $200

Prepaid balance

Top up in fixed amounts from $5 to $1,000 and spend it down request by request.

$5$100$1,000

Start saving 20–70% today

Create a key in under a minute, top up, and ship. Only pay for what you use.

Get your API keyRead the docs