Documentation

Quickstart

Create an account, copy the key from the dashboard, and point a client at the base URL.

Base URL

OpenAI SDK (Python)

from openai import OpenAI
client = OpenAI(base_url="https://voidox.cc/v1", api_key="sk-voidox-...")
client.chat.completions.create(model="MODEL_ID", messages=[{"role": "user", "content": "hi"}])

Anthropic SDK (Python)

from anthropic import Anthropic
client = Anthropic(base_url="https://voidox.cc", api_key="sk-voidox-...")
client.messages.create(model="MODEL_ID", max_tokens=256, messages=[{"role": "user", "content": "hi"}])

Claude Code

export ANTHROPIC_BASE_URL="https://voidox.cc"
export ANTHROPIC_API_KEY="sk-voidox-..."
export ANTHROPIC_MODEL="gemini-flash-3.8"

Cursor, Cline or any OpenAI-compatible client

Base URL https://voidox.cc/v1, the key from your dashboard, and a model id from the table below.

Endpoints

• POST /v1/chat/completions — OpenAI shape, streaming and JSON
• POST /v1/messages — Anthropic shape, streaming and JSON
• GET /v1/models — OpenAI list envelope; validates a presented key

Send the key as Authorization: Bearer sk-voidox-... or as x-api-key.

Models and pricing

Live from the registry. Prices are per 1M tokens. You are billed on the prompt and completion tokens actually used.

Errors

Errors come back in the shape of the surface you called.

OpenAI surface

{"error":{"message":"...","type":"authentication_error","param":null,"code":"401"}}

Anthropic surface

{"type":"error","error":{"type":"api_error","message":"..."}}

• 401 missing, invalid or revoked key
• 402 balance cannot cover the request
• 404 unknown or disabled model
• 413 request body too large
• 429 per-key rate limit, honour Retry-After
• 502 upstream unavailable

Limits

• Request body up to 32 MB, so images and long prompts fit.
• Per-key rate limit with a burst allowance; a 429 carries Retry-After.
• max_tokens above 8192 is clamped for the pre-authorization check.
• Balance is checked before any upstream call, so a request either fits or is refused.