A lot of LLM proxies quietly add a margin on top of provider pricing — sometimes 10-20%, rarely disclosed clearly. When we built RouterPlex, we made a deliberate call to not do that.
How it works
Each of the 28 models we route to (Claude, GPT, DeepSeek, Qwen, MiniMax, and others) is billed to you at the same per-token rate the vendor publishes on their own pricing page. We keep a public pricing sheet so you can check it yourself rather than take our word for it.
Why this matters
If you're building a product on top of an LLM API, your margins are already thin enough without an invisible reseller markup baked into your infra costs. Transparent pricing also means our incentives are aligned with yours — we don't make more money by routing you to a more expensive model than you need.
The other half: payment access
The other problem we wanted to solve: getting a card that OpenAI or Anthropic billing will accept isn't guaranteed in every market. RouterPlex accepts crypto top-ups (USDT) with no KYC, in addition to standard card payment.
Try it
One OpenAI/Anthropic-compatible key, 28 models, works natively with Claude Code and VS Code's built-in chat.
Feedback (especially on pricing transparency — did we miss anything?) genuinely welcome.
Top comments (0)