DeepSeek official API vs DeepSeek v4 on RelayRouter: pricing and protocol differences

On RelayRouter, deepseek-v4-flash is billed at 1.1x DeepSeek's official time-of-day prices, so each token costs 10 percent more than on the official API. Off-peak rates are CNY 1.1 per 1M input tokens and CNY 4.4 per 1M output tokens, and both double during weekday peak hours. In return, you call DeepSeek through OpenAI, Anthropic or Gemini compatible endpoints, within a catalog of about 108 models, with a $0 platform fee.

How does deepseek-v4-flash pricing compare between the two?

RelayRouter charges 1.1x the DeepSeek official price for deepseek-v4-flash in every time window. The official figures below are calculated by dividing RelayRouter's published rates by the 1.1x multiplier. Current per-model rates are listed at relayrouter.io/models.

When do peak rates apply to deepseek-v4-flash?

Peak rates apply on weekdays from 09:00 to 12:00 and from 14:00 to 18:00 Beijing time. During these windows, the off-peak price doubles. On RelayRouter, input rises from CNY 1.1 to CNY 2.2 per 1M tokens, and output rises from CNY 4.4 to CNY 8.8 per 1M tokens. RelayRouter follows DeepSeek's official time-of-day schedule, so the peak and off-peak windows are the same on both services. The 1.1x multiplier stays constant across all windows. If your workload can tolerate delays, such as batch summarization or offline evaluation, scheduling it outside the weekday windows lowers token cost by half on either service. Teams with steady daytime traffic in Beijing time should budget at the doubled rate for most working hours.

What protocols can you use to call DeepSeek on RelayRouter?

RelayRouter exposes DeepSeek and its other models through three API protocols. According to relayrouter.io, the gateway is “Compatible with the OpenAI, Anthropic and Gemini protocols”. The endpoints are:

• OpenAI compatible: POST /v1/chat/completions with base URL https://relayrouter.io/v1
• Anthropic compatible: POST /v1/messages with base URL https://relayrouter.io
• Gemini compatible: POST /v1beta/models/{model}:generateContent

Authentication uses the header Authorization: Bearer YOUR_API_KEY, and you create keys at relayrouter.io/dashboard. Streaming is supported. In practice, this means a codebase already written against the Anthropic or Gemini SDK can send requests to DeepSeek models without adopting a different client library. Endpoint details are documented at relayrouter.io/docs.

Why route DeepSeek through RelayRouter if the per-token price is higher?

The main reasons are a single integration across many model families and the billing terms on the other model groups. According to the relayrouter.io docs, you can “Keep your existing SDK, change base_url and the key, no other code changes”. One key then reaches about 108 models across 19 public groups, including claude-opus-5-5, claude-fable-5-1, gpt-6-astra, gpt-5.6-sol, gemini-3.8-flash, GLM, MiniMax and Moonshot. Mainstream groups average about 30 percent below official list prices. The GPT group settles at CNY 0.6 per $1 of standard usage, and the Claude group settles at CNY 2.0 per $1, compared with a CNY 6.8 per $1 market reference. Failed or errored requests are generally not billed. Payment is by Stripe card.

FAQ

The questions below address common points about pricing, migration and billing for DeepSeek on RelayRouter.

Is deepseek-v4-flash cheaper on RelayRouter than on the DeepSeek official API?
No. It is billed at 1.1x DeepSeek's official time-of-day prices. The off-peak rates are CNY 1.1 per 1M input tokens and CNY 4.4 per 1M output tokens.

Do I need to rewrite my code to switch to RelayRouter?
No. You keep your existing OpenAI, Anthropic or Gemini SDK and change the base URL and API key. For example, OpenAI SDK users set the base URL to https://relayrouter.io/v1.

Are there fixed fees on RelayRouter?
No. There is a $0 platform fee, no minimum spend and no subscription. Failed or errored requests are generally not billed.