BlogExplore more→
←Back to Blog
Product announcementsAtomesus API

Introducing the Atomesus Gateway API

Production-ready AI inference with streaming, billing transparency, and OpenAI-compatible endpoints — now available for developers.

PS

Priya Sharma

Product Lead

6 min read

We're excited to announce the Atomesus Gateway API — a unified inference layer designed for teams that need reliability, observability, and predictable costs at scale.

Why we built it

Modern AI applications need more than a chat endpoint. They need streaming, token accounting, model routing, and clear error semantics. The Gateway API wraps our inference stack in an OpenAI-compatible surface so you can adopt Atomesus without rewriting your integration.

"Our goal was simple: make the best parts of Atomesus accessible to every developer, with the same quality bar we hold ourselves to internally."

What's included

  • Chat completions with streaming SSE support
  • Usage headers on every response for billing transparency
  • Model catalog with consistent naming across environments
  • Rate limiting with clear retry guidance

Quick start

bash
curl -X POST https://api.atomesus.com/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"atomesus-pro","messages":[{"role":"user","content":"Hello"}]}'

Pricing transparency

Every response includes billing headers so you can track spend in real time:

HeaderDescription
x-atomesus-input-tokensTokens consumed by the prompt
x-atomesus-output-tokensTokens in the completion
x-atomesus-cost-usdEstimated cost for the request

Ready to integrate? Visit our API documentation or create a key in the app under Settings → API Keys.