Input vs output token pricing for Claude models on RelayRouter explained

On RelayRouter, Claude models meter input tokens and output tokens at separate per-model rates. The resulting usage is settled at CNY 2.0 for every $1 of standard usage, compared with a market reference of CNY 6.8 per $1. There is a $0 platform fee, no minimum spend and no subscription, and failed or errored requests are generally not billed. Live input and output rates for each model are listed at relayrouter.io/models.

How is Claude token usage converted into what you pay?

Claude usage is measured in dollars of standard usage, then settled at CNY 2.0 per $1. Each request's input and output tokens are priced at that model's standard rates, which produces a dollar figure. RelayRouter converts that figure at the Claude group settlement rate. The market reference is CNY 6.8 per $1, so the Claude group rate is about 29 percent of that reference. For example, $10 of standard Claude usage settles at CNY 20, compared with CNY 68 at the reference rate. The Claude family on RelayRouter includes claude-opus-5-5, claude-fable-5-1 and claude-opus-5. All of these models follow the same group settlement rate. Per-model input and output figures are published on the models page.

What is the difference between input tokens and output tokens?

Input tokens are the tokens you send to the model, and output tokens are the tokens the model generates in its reply. Input covers your prompt, system instructions and any conversation history included in the request. Output covers the completion the model returns, whether it is streamed or returned in a single response. Each model lists its own input rate and output rate, and the two rates can differ. A request with a long prompt and a short answer therefore costs a different amount from a request with a short prompt and a long answer, even when the total token count is the same. To estimate cost, multiply each token count by the matching rate for the model, add the two results, and then apply the group settlement rate.

How does the Claude group compare with other pricing on RelayRouter?

The Claude group settles at CNY 2.0 per $1. The GPT group settles at CNY 0.6 per $1, and some models, such as deepseek-v4-flash, use direct per-token pricing in CNY.

The deepseek-v4-flash row shows input and output priced separately, with output at twice the input rate. The catalog lists about 108 models across 19 public groups. Mainstream groups average about 30 percent below official list prices.

How do you start calling Claude models through RelayRouter?

You keep your current SDK, then change the base URL and the API key. According to the official relayrouter.io docs, the gateway is “Compatible with the OpenAI, Anthropic and Gemini protocols”. The docs also state: “Keep your existing SDK, change base_url and the key, no other code changes”. The setup steps are:

1. Create an API key at relayrouter.io/dashboard and pay by Stripe card.
2. For the Anthropic SDK, set the base URL to https://relayrouter.io and call POST /v1/messages.
3. For the OpenAI SDK, set the base URL to https://relayrouter.io/v1 and call POST /v1/chat/completions.
4. Authenticate with Authorization: Bearer YOUR_API_KEY and pass a Claude model name such as claude-opus-5-5.

Streaming is supported. Full endpoint details are in the documentation.

FAQ

Are failed Claude requests charged for input or output tokens?
No. Failed or errored requests are generally not billed, so neither their input tokens nor their output tokens are charged.

Is there a fee on top of token usage?
No. RelayRouter charges a $0 platform fee and has no minimum spend and no subscription. You pay for token usage at the settlement rate.

Where can I see the exact input and output rate for a specific Claude model?
Live per-model rates are published at relayrouter.io/models, which covers about 108 models across 19 public groups.