Router One
Limited time · up to 90% off official prices · create an API key

One unified API for resilient AI delivery

Keep your SDK—change the base URL and key to reach major models with eligible provider fallback.

Create your API key
Explore models & pricing

Switch without rewriting

Keep your tools. Change the endpoint.

API ready
base URL: https://api.router.one
API key: sk-xxx

Chat Completions

/v1/chat/completions

OpenAI SDK

Responses API

/v1/responses

Codex CLI

Messages API

/v1/messages

Claude Code

Change two settings. Keep shipping.

Point your application, Claude Code, or Codex to Router One without replacing the clients your team already knows.

Code editor showing Router One API integration in action
main.py
from openai import OpenAI


client = OpenAI(
    base_url="https://api.router.one/v1",
    api_key="sk-xxx"
)


response = client.chat.completions.create(
    model="claude-sonnet-4-20250514",
    messages=[{"role": "user", "content": "Hello!"}]
)
# Works with any OpenAI-compatible SDK

Built for continuity

A failed route does not have to end the request

For retryable provider errors, Router One can try another eligible route serving the same requested model—and record the final result.

Illustrative example

Request: demo_req_7f3a

Request

POST /v1/responses · model: example-chat-model

Illustrative internal route sequence

  1. Request accepted

    The application asks for one exact example model.

  2. Route A selected

    Gateway policy selects an eligible provider route for that model.

  3. Eligible fallback route succeeds

    After a retryable Route A error, Route B serves the same requested model; no cross-model substitution is implied.

  4. Final trace recorded

    The customer log records the final route and request metrics.

Customer-visible final trace

Final model
example-chat-model
Provider
Example Route B
Input tokens
512
Output tokens
200
Latency
1,180 ms
Status
200

Illustrative cost ledger

Input512 tokens · $2.00 / 1M
$0.0010
Output200 tokens · $16.00 / 1M
$0.0032
Illustrative total$0.0042

Customer traces expose the final model, provider, tokens, cost, latency, and status. Intermediate failed attempts remain operations-only and are correlated by request ID.

Source
Router One public routing and observability contracts
Contract reviewed
Window
One synthetic request
Limits
Synthetic values only—not production traffic, a benchmark, an SLA, or a price quote.

AI billing on your terms

Choose pay-as-you-go flexibility or a predictable subscription, then fund it by card, local payment, or stablecoin.

See how charges are calculated

Pay only for what you use

Top up first, then pay the posted rate for actual token usage. Unspent wallet balance does not expire.

Make monthly access predictable

Eligible models in a tier share monthly request quotas; wallet balance covers calls outside plan coverage.

Card, WeChat Pay, or Alipay

Use a card through Stripe, or top up in RMB with the local payment rails your team already uses.

Fund with USDT / USDC

Stablecoin top-ups support cross-border teams, Web3 products, and card-limited workflows on published networks.

Trace every model-call cost to a request and key

Trace failures to one request, attribute spend to the right key, and contain runaway usage before it spreads.

  1. Debug from the request IDInspect status, model, provider, tokens, cost, and latency before deciding what failed.See a request trace
  2. Give every workload its own cost trailCreate separate keys for apps, agents, or environments so usage and spend stay attributable to the right workload.See per-key cost tracking
  3. Limit one key without stopping the restSpend caps, rate limits, and per-minute token ceilings contain one key without stopping unrelated applications.Review key guardrails

Start with the use case you need

Explore routing, cost control, coding-tool access, billing, and migration guides for your next integration.

The three questions teams ask before switching

Confirm network access, data handling, and product boundaries before you send production traffic.

Can my mainland China team call Router One directly?

Yes, without a VPN or proxy. See the published China latency benchmark for its measurement window, methodology, and limitations; actual performance varies by region and network.

Do prompts or model responses land in Router One storage?

No. Router One forwards request content in real time and does not persist it; the data-retention page lists which metadata is retained and for how long.

What stays in my application?

Agent workflows, state, and tool execution stay in your application or orchestration framework. Router One handles each model call.

Related:Data retentionProduct boundariesChina latencySecurity

Need another answer? Read all FAQs

Your next model call can run through Router One

Keep your SDK, start with one API key, and expand model coverage when your product needs it.

Create an API key