RE — Unified API Key Pool

Why Use RE?

Auto-Rotating Pool

Add multiple credentials per provider. The pool picks a working key at random for each request. If one fails, it tries the next — no manual intervention needed.

Per-Key Validation

Every key is tested before it enters the pool. Expired, malformed, or revoked tokens are rejected immediately. You only add keys that actually work.

Sliding-Window Rate Limits

Limits scale automatically with your active providers. More keys = higher RPM. Usage resets every minute and every day — no sudden caps.

Automatic Alerts

When a credential stops working, the system disables it and sends a notification. You know exactly which key failed and why — no silent outages.

Single API Key

One API key for every provider. Point your client to one endpoint with one Authorization header. No more managing multiple tokens across services.

OpenAI-Compliant

Drop-in replacement for any OpenAI SDK. Just change the base URL and API key. Streaming, function calling, reasoning — all supported through the standard format.

Supported Providers

Kimi

moonshot

k2d6k2d6-thinkingk3

DeepSeek

deepseek

deepseek-v4-flashdeepseek-v4-pro

Qwen

Alibaba

qwen3.8-max-previewqwen3.7-plusqwen3.7-maxqwen3.6-plusqwen3.6-max-previewqwen3.6-27bqwen3.6-35b-a3bqwen3.5-plusqwen3.5-flashqwen3.5-397b-a17bqwen3.5-omni-plusqwen3.5-omni-flashqwen3-max-2026-01-23qwen-plus-2025-07-28qwen3-coder-plusqwen3-vl-plusqwen3-omni-flash-2025-12-01

Gemini

Google

gemini-3-flashgemini-3-flash-extendedgemini-3-progemini-3-pro-extendedgemini-3-flash-litegemini-3-flash-lite-extendedgemini-3-flash-thinking

How It Works

Create an Account

Add Credentials

Paste your provider API keys — each key is validated before being accepted into the pool.

Use a Single Endpoint

Point your client to /v1/chat/completions with your API key. The pool handles the rest.

Quick Start

cURL Example

# Replace <your-api-key> with the key from your dashboard
curl -X POST http://re.sixfinger.live/v1/chat/completions \
  -H "Authorization: Bearer <your-api-key>" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "k2d6",
    "messages": [{"role": "user", "content": "hello"}]
  }'

Frequently Asked Questions

What providers are supported?

Kimi, DeepSeek, Qwen, and Gemini. Each provider appears as its own model family — just set the model parameter to k2d6, deepseek-v4-flash, qwen3.5-flash, or gemini-3.5-flash in your request.

How do I get credentials?

Open each provider's web interface, open the browser console (F12), and copy the token from local storage or network requests. Each provider has a step-by-step guide in the Credentials page after you sign in.

What happens when a key stops working?

After 3 consecutive errors, the key is automatically disabled and removed from the pool. A notification is created in your dashboard so you know exactly which key failed. You can re-enable it by re-adding fresh credentials.

How do rate limits work?

Your rate limit depends on how many different provider keys you have active. 0 active providers = 5 RPM, 1 = 15 RPM, 2 = 30 RPM, 3+ = 100+ RPM. Each minute and each day the counter resets. You can monitor your usage in the API Usage tab.

Is this production-safe?

This is a reverse-engineered proxy for free-tier web interfaces, not an official API. Credentials are session-based and expire. Use a secondary or disposable account. The system is intended for personal use and experimentation.

One endpoint to pool every free AI token