DEV Community

RouterPlex
RouterPlex

Posted on

RouterPlex: One API Key for 28 LLMs — Claude, GPT, DeepSeek, Qwen, MiniMax

#ai #api #llm #startup

Most projects that touch multiple LLM providers end up with a pile of vendor SDKs, a pile of API keys, and separate billing relationships to manage. RouterPlex is a gateway that collapses that down to one key.

What it does

One OpenAI- and Anthropic-compatible API key routes to 28 models across providers — Claude, GPT, DeepSeek, Qwen, MiniMax, and more. If your tooling already speaks the OpenAI or Anthropic /v1/messages format, you don't rewrite anything:

curl https://dub.sh/no1oLNt \
  -H "Authorization: Bearer $ROUTERPLEX_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet-4-6",
    "messages": [{"role": "user", "content": "Hello"}]
  }'
Enter fullscreen mode Exit fullscreen mode

It also works as a drop-in for Claude Code and VS Code's built-in AI chat — point the base URL at RouterPlex and pick any of the 28 models.

Why we built it

Two problems kept showing up for developers outside a handful of markets:

  1. Billing access. Getting a card that OpenAI's or Anthropic's billing will actually accept isn't guaranteed everywhere. RouterPlex accepts crypto top-ups (USDT) with no KYC, alongside normal card payment.
  2. Pricing trust. A lot of LLM proxies quietly mark up provider pricing. RouterPlex bills at the official vendor list price — you can check our sheet against the provider's own pricing page.

Where it's at

Early — a handful of real users so far, still finding rough edges. If you're building anything that touches multiple LLM providers, I'd genuinely appreciate you trying it and telling me what's broken or missing.

https://dub.sh/CZbfvUc

Top comments (0)