TokioAI LLM Router

One API, 20+ models. Drop-in OpenAI replacement with built-in PII protection and prompt injection firewall.

OpenAI Compatible PII Auto-Redaction Prompt Firewall Usage Tracking
Auth (JWT / API Key)
Rate Limiter
Prompt Firewall
PII Sanitizer
LLM Provider
Usage Tracker
🔌

Drop-in OpenAI Replacement

Same /v1/chat/completions endpoint. Change your base URL and you are done. Works with any OpenAI SDK or library.

🛡

PII Auto-Redaction

Credit cards, SSNs, emails, phone numbers, addresses -- automatically stripped before reaching any LLM provider. Your data never leaves your control.

🔥

Prompt Injection Firewall

Multi-layer detection blocks jailbreaks, role hijacking, encoding tricks, and system prompt extraction attempts in real-time.

📊

Usage Tracking & Billing

Per-request token counting, cost tracking by model, daily/monthly usage dashboards. Know exactly what you are spending.

Supported Models

GPT-4o
OpenAI
GPT-5
OpenAI
o3 / o3-mini
OpenAI
Claude Opus 4
Anthropic
Claude Sonnet 4
Anthropic
Gemini 2.5 Pro
Google
Gemini 2.5 Flash
Google
DeepSeek R1
DeepSeek
Kimi K2.7 Code
Moonshot
Kimi K3
Moonshot
DeepSeek Chat
DeepSeek
+ 15 more via OpenRouter
Any model

Quick Start

# Just change your base URL
from openai import OpenAI

client = OpenAI(
    base_url="https://router.tokioia.com/v1",
    api_key="tr_your_api_key_here",
)

response = client.chat.completions.create(
    model="claude-sonnet-4",  # or gpt-4o, gemini-2.5-flash, kimi...
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

# Works with curl too:
# curl https://router.tokioia.com/v1/chat/completions \
#   -H "Authorization: Bearer tr_your_key" \
#   -d '{"model":"gpt-4o","messages":[{"role":"user","content":"Hi"}]}'

Get Started

Create your account and get an API key in seconds.

Create Account API Documentation