The AI API for your team.

An OpenAI-compatible endpoint for your developers. Issue API keys, track usage and cost per user, and manage it all from one console.

# Works with the official OpenAI SDK — just change the base URL
from openai import OpenAI

client = OpenAI(
    base_url="http://c9jm29xefr1ye3myebf4q9r6.lata.ethans.app/v1",
    api_key="sk-spark-...",
)
resp = client.chat.completions.create(
    model="sparkai-default",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(resp.choices[0].message.content)

OpenAI-compatible

Point any OpenAI client at /v1. Chat completions, streaming, and legacy completions all work out of the box.

Streaming built in

Token-by-token streaming over server-sent events, just like the OpenAI API, plus standard non-streaming responses.

Configured by your admins

Your team's admins manage the platform from one console, so developers can focus on building.

Cost & usage analytics

Every request is metered — tokens, compute time, and dollars, broken down per user and per day.

Self-serve keys

Users mint and revoke their own API keys and read the docs without bugging you.

Admin control

Provision accounts, set pricing, flip the kill switch, and watch spend from one console.