Quickstart

About two minutes, assuming you already have Python or Node installed.

1. Get a key

Sign in to the console, fund a balance, then create an API key. The secret is shown once. We store only a hash of it, so if you lose it we have no way to show it to you again — create a new key and revoke the old one.

Keep it out of client-side code and out of git. Anything that can read the key can spend your balance.

2. Point your client at the gateway

African Open Router speaks OpenAI's chat-completions API. If your code already talks to OpenAI, the only change is the base URL and the key.

# Python
from openai import OpenAI

client = OpenAI(
    base_url="https://api.redactron.ai/v1",
    api_key="<AOR_API_KEY>",
)

// Node
import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.redactron.ai/v1",
  apiKey: process.env.AOR_API_KEY,
});

3. Send a request

response = client.chat.completions.create(
    model="gpt-5.6-sol",
    messages=[{"role": "user", "content": "Say hello in three words."}],
)
print(response.choices[0].message.content)

Or with curl:

curl https://api.redactron.ai/v1/chat/completions \
  -H "Authorization: Bearer $AOR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.6-sol","messages":[{"role":"user","content":"Hello"}]}'

4. Streaming

Pass stream=True and read chunks as they arrive. A stream that fails midway is reported as a failed request, not as a short successful one, so you never have to guess whether the text you got is complete.

stream = client.chat.completions.create(
    model="gpt-5.6-sol",
    messages=[{"role": "user", "content": "Count to five."}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="", flush=True)

Models

Use the provider's own model name — gpt-5.6-sol — not an African Open Router alias. The console catalog is the reference list of what you can call and what it costs; the pricing page shows current rates.

Errors you will actually hit

Status	Code	What to do
401	invalid_token	The key is wrong, revoked, or not published yet. Check that you sent `Authorization: Bearer <key>`.
400	invalid_request	The body failed schema validation — usually a missing `model`, an empty messages list, or a model name with stray whitespace in it.
402	wallet_denied	Your balance does not cover the request's worst-case cost. Top it up.
404	model_unavailable	The model is not in your workspace's catalog.
429	—	Rate limited at the edge, before the gateway ever sees the request. This one is not JSON: the body is `error code: 1015` as `text/plain`, and the wait is in the `Retry-After` header, in seconds. Detect 429 from the status code, not from the body — every other row here uses the JSON envelope above.
503	provider_unavailable	The upstream provider failed. Retrying is safe — you are not charged for a request that produced nothing.

What a request costs you

Before we call the provider we reserve the request's worst-case cost against your balance. When it finishes we settle the real amount and return the difference. A request refused before it reaches a provider costs nothing. Every settled request appears in the console with its token counts and its exact charge.

Rotating a key

Create the new key first, deploy it, then revoke the old one — in that order, so nothing is ever running for even a moment without a valid credential.

Getting help

[email protected]. Include the request_id from the error body; it is the fastest way for us to find out what happened to one specific request.