Dev Resources
XiuRouter favicon

XiuRouter

AI/API Building/Tooling

One API for GPT, Claude, and Gemini with native protocol routes, scoped keys, usage records, and transparent usage-based pricing.

XiuRouter's website screenshot

About XiuRouter

XiuRouter is an API routing service that provides a single interface for accessing multiple AI model providers, including Claude, GPT, Gemini, and Grok models. It supports OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, and Gemini GenerateContent protocols, allowing clients and SDKs to connect using their native request format by pointing to XiuRouter's base URL with an API key.

The service organizes model access into service-source groups (such as "Discount" and "Full-access") that offer different prices for the same models, with per-model input and output token pricing compared against official provider rates. Users top up a wallet balance, create API keys scoped to specific models and groups, and are billed based on actual token usage, with a dashboard for reviewing usage records, request status, latency, and model call rankings.

XiuRouter is operated by XiuLab Inc, a U.S. corporation, and is one of several products offered alongside XiuStore, XiuBench, and XiuGEO. It is aimed at developers and teams building applications or agents that need programmatic access to multiple large language model providers through a unified API.

Key features

  • Single API for Claude, GPT, Gemini and Grok models with up to 90%+ savings
  • Supports OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, and Gemini GenerateContent protocols
  • Multiple service-source groups (e.g. Discount, Full-access) with different prices per model
  • Built-in playground to send test requests and inspect streaming responses
  • Usage records showing model, group, token usage, cost, request status and latency by API key
  • Model call rankings showing usage share over the last 7 days

Frequently asked questions

Which API protocols are supported?

XiuRouter supports OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, and Gemini GenerateContent. You use the Base URL and API key matching your tool's protocol.

How much can I save on model pricing?

Savings vary by model and group, with some models seeing over 90% off official input or output prices; for example gpt-5.6-sol input is discounted 94.2%. The 90%+ maximum does not apply to every model, group, or output price.

Which models are supported?

The current catalog includes leading Claude, GPT, Gemini and Grok models, and it changes with upstream availability. The pricing page is the authoritative source for the current model list.

How do service-source groups work?

Groups represent different service sources and prices, such as Discount and Full-access, and available models are shown on the pricing page and when creating an API key. A group name does not promise a particular speed, quality, or service level.

How is billing calculated?

Billing is based on the model, service-source group, and actual input, output, and cache token usage, deducted from a prepaid wallet balance. The pricing page lists per-1M-token prices, while usage records show each request's settlement.

Ad
SerpApi - Search API favicon

SerpApi - Search API

SerpApi's Search API makes it easy and fast to scrape Google and other search engines.

Learn more

Featured products

  • SerpApi - Search API favicon
  • TranscriptFetch favicon

Show your product to thousands of developers

· 100k monthly pageviews
· 7k newsletter subscribers

Advertise your product