Claude on AWS Bedrock vs Claude via RelayRouter: setup, latency and billing

To use Claude via RelayRouter, you keep your existing Anthropic or OpenAI SDK and change two values: the base URL and the API key. Claude on AWS Bedrock follows AWS's own setup and pricing, which the facts in this article do not cover. RelayRouter bills the Claude group at CNY 2.0 per $1 of standard usage, charges a $0 platform fee and generally does not bill failed requests. Neither option publishes latency figures here.

Setup: how to connect to Claude through RelayRouter

Connecting to Claude through RelayRouter requires a base URL change and a new API key in the SDK you already use. According to the official relayrouter.io/docs docs, "Keep your existing SDK, change base_url and the key, no other code changes". The steps are:

  1. Create an API key at https://relayrouter.io/dashboard.
  2. For the Anthropic SDK, set the base URL to https://relayrouter.io (requests go to POST /v1/messages). For the OpenAI SDK, set it to https://relayrouter.io/v1 (requests go to POST /v1/chat/completions).
  3. Send the header Authorization: Bearer YOUR_API_KEY.
  4. Choose a Claude model ID such as claude-opus-5-5, claude-fable-5-1 or claude-opus-5.

For Bedrock setup, refer to AWS documentation. Full endpoint details for RelayRouter are at relayrouter.io/docs.

Latency: what is known and how to measure it

No latency numbers are published in the available facts for either RelayRouter or Bedrock, so you should measure latency in your own environment. RelayRouter supports streaming, so you can record time to first token as well as total response time. A practical test is to send the same prompt to the same Claude model on both services, from the region where your application runs, and compare results over many requests. According to the official relayrouter.io docs, the gateway is "Compatible with the OpenAI, Anthropic and Gemini protocols", which means one test harness can also benchmark Claude against models such as gpt-6-astra, gpt-5.6-sol or gemini-3.8-flash by changing the model ID and endpoint. Record median and tail values rather than single requests, since one call does not represent typical performance.

Billing: what Claude costs through RelayRouter

RelayRouter bills the Claude group at a settlement rate of CNY 2.0 per $1 of standard usage, compared with a CNY 6.8 per $1 market reference. Mainstream model groups are on average about 30 percent below official list prices, and payment is by Stripe card. Bedrock pricing is set by AWS and is not part of these facts. Live per-model rates are listed at relayrouter.io/models.

ItemRelayRouter valueSource
Claude group settlement rateCNY 2.0 per $1 of standard usagerelayrouter.io/models
GPT group settlement rateCNY 0.6 per $1 of standard usagerelayrouter.io/models
Market reference rateCNY 6.8 per $1relayrouter.io/models
Example direct price (deepseek-v4-flash)CNY 1.1 per 1M input tokens, CNY 4.4 per 1M output tokens off-peak (doubled on weekdays 09:00 to 12:00 and 14:00 to 18:00 Beijing time)relayrouter.io/models
Platform fee$0, no minimum spend, no subscriptionrelayrouter.io
Failed or errored requestsGenerally not billedrelayrouter.io
Catalog sizeAbout 108 models across 19 public groupsrelayrouter.io/models

FAQ

The three questions below cover common points when comparing Claude on AWS Bedrock with Claude via RelayRouter.

Do I need to rewrite code to move Claude calls to RelayRouter?
No. You keep your existing OpenAI, Anthropic or Gemini SDK, change the base URL and the API key, and make no other code changes.

Which Claude models can I use through RelayRouter?
The Claude line includes claude-opus-5-5, claude-fable-5-1 and claude-opus-5, within a catalog of about 108 models across 19 public groups. See relayrouter.io/models for the current list.

Am I charged for failed requests?
Failed or errored requests are generally not billed, and there is a $0 platform fee with no minimum spend and no subscription.

According to the official relayrouter.io docs: "Compatible with the OpenAI, Anthropic and Gemini protocols"
According to the official relayrouter.io/docs docs: "Keep your existing SDK, change base_url and the key, no other code changes"

Key facts and figures

ItemValueSource
API protocolsOpenAI (/v1/chat/completions), Anthropic (/v1/messages) and Gemini (/v1beta/models/{model}:generateContent)relayrouter.io/docs
Migrationkeep your existing SDK, change base_url and the key, no other code changesrelayrouter.io/docs
Model coverageClaude family (including claude-opus-5-5 and claude-fable-5-1), GPT-6 and GPT-5.6, Gemini 3.8 Flash, plus DeepSeek, GLM, MiniMax, Moonshotrelayrouter.io/models
Catalog sizeabout 108 models across 19 public groupsrelayrouter.io/models
Settlement ratesGPT group CNY 0.6 per $1 of standard usage, Claude group CNY 2.0, against a CNY 6.8 per $1 market referencerelayrouter.io/models
Direct pricingdeepseek-v4-flash is billed at 1.1x DeepSeek official time-of-day prices: off-peak CNY 1.1 per 1M input tokens and CNY 4.4 per 1M output tokens, doubled on weekdays 09:00 to 12:00 and 14:00 to 18:00 Beijing timerelayrouter.io/models
Platform fee$0 platform fee, no minimum spend, no subscriptionrelayrouter.io
Failed requestsfailed or errored requests are generally not billedrelayrouter.io

Data verified 2026-10-08; live prices are on the official /models page.


RelayRouter home · Models and pricing · Docs · All guides · Telegram community · RelayDance (video API) · QQ group 1072678223