An OpenAI-compatible API, paid with QRIS
Buy credit, get a key, start calling. No sign-up, no subscription — credit does not expire.
Payments are temporarily unavailable.
We are restoring the payment connection — please check back shortly.
Existing API keys keep working normally.
Pricing
| Tier | Input / 1M tokens | Output / 1M tokens |
|---|---|---|
| Lite Fast small models for high-volume, latency-sensitive work |
$0.15 | $0.60 |
| Standard Balanced quality and cost for general use |
$0.50 | $1.20 |
| Pro Largest models, highest quality |
$1.20 | $3.00 |
You are charged per token actually used. Failed requests cost nothing.
Models
| Model | Tier | Context |
|---|---|---|
| gpt-oss-20b | lite | 32,768 |
| phi-4 | lite | 32,768 |
| qwen3-14b | lite | 32,768 |
| qwen3-8b | lite | 32,768 |
| gpt-oss-120b | pro | 32,768 |
| qwen3-235b-a22b | pro | 32,768 |
| glm-4-32b | standard | 32,768 |
| mistral-small-3.1-24b | standard | 32,768 |
| qwen3-30b-a3b-instruct | standard | 32,768 |
| qwen3-32b | standard | 32,768 |
| qwen3-coder-30b-a3b-instruct | standard | 32,768 |
How to use it
from openai import OpenAI
client = OpenAI(
api_key="rbl_live_...",
base_url="https://yolahproxy.com/v1",
)
response = client.chat.completions.create(
model="qwen3-32b",
messages=[{"role": "user", "content": "Hello"}],
)
Any OpenAI-compatible client works — just change the base URL.