golin
Sign inGet API key
Pay-as-you-go AI credits

One AI endpoint.
Billed by the token.

Buy AI API access from 3× wholesale and drop it into any project. We handle the supplier relationship, metering, and billing so you can ship.

Get your API keyRead the docs
Markup
3×
Precision
1e-4
Servers
Zero
python
from openai import OpenAI

client = OpenAI(
    api_key="sk-your-golin-key",
    base_url="https://api.roma.co.ke/v1",
)

stream = client.chat.completions.create(
    model="claude-3-5-sonnet",
    messages=[{"role": "user", "content": "Write a haiku about billing."}],
    stream=True,
)

for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")
bash
curl https://api.roma.co.ke/v1/chat/completions \
  -H "Authorization: Bearer sk-your-golin-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Everything between you and the model

Auth, metering, billing, audit, and abuse control — running at the edge with no servers to operate.

Drop-in endpoint
Point any [OI]-compatible SDK at our base URL. No proxy code, no SDK rewrite, no schema translation.
Prepaid credits
Top up in KES with M-Pesa or card. Balance is deducted per token and never goes negative.
Sub-cent accounting
Usage is stored in microdollars (USD × 10,000) so a $0.0001 request never rounds to zero.
Live usage log
Every request is recorded with tokens, model, cost, and the balance that followed it.
Bot and abuse control
Turnstile on registration, suspension controls, and a full activity trail on every action.
Rotatable keys
Rotate an API key at any time from the dashboard. The old key stops working immediately.

Three steps to production

  1. 1

    Create an account

    Register with email or Google and get an API key shown exactly once.

  2. 2

    Add credits

    Pay in KES through Paystack. Credits settle to your balance instantly.

  3. 3

    Call the endpoint

    Use the [OI] SDK you already have. We meter, bill, and log every request.

Transparent per-token pricing

A flat 3× on wholesale token cost. No subscription, no seat fees, no minimum spend.

Add credits
Token typeFieldYou pay / 1MEffective markup
Inputprompt_tokens$0.01620067%
Outputcompletion_tokens$0.06480067%
Cache readcache_read_input_tokens$0.00032467%
Cache writecache_creation_input_tokens$0.01620067%

Completion tokens already include reasoning/thinking tokens. Balances are stored in microdollars and rounded up to the nearest 1/10,000 of a dollar.

One base URL

Bring the model you already use

Model availability follows your supplier plan. If it is exposed upstream, it is callable through the same endpoint, billed identically.

ClaudeGPTGeminiLlamaMistralDeepSeekGrokQwen

Run GET /v1/models with your key for the live list.

Compatible by design
Standard request and response shapes, including streaming and cache-aware usage.

Bearer or x-api-key

Authenticate with whichever header your client already sends.

Streaming passthrough

Server-sent events are forwarded as they arrive, then billed from the final usage chunk.

Itemised usage

Prompt, completion, cache-read and cache-write tokens per request.

Frequently asked

Still stuck? Read the docs.

Get an API key in under a minute

Register, top up, and call the endpoint. Your key is shown once, so keep it safe.

Create accountSign in