# Router One

Unified LLM API gateway with smart routing, fallback, and cost observability

> Source: [Pro Launch](https://prolaunch.net) — a launch platform for startups and new tools.

- Pro Launch page: https://prolaunch.net/projects/router-one
- Markdown version: https://prolaunch.net/projects/router-one.md
- Website: https://router.one/?ref=4E2F975D663212E2
- Categories: Artificial Intelligence & ML, APIs & SDKs, SaaS
- Launch status: launched
- Launch date: 2026-09-29
- Upvotes: 3

## About

Router One is a unified LLM API gateway that provides a single OpenAI-compatible endpoint (https://api.router.one/v1) for accessing models from OpenAI, Anthropic, Google, Grok, and more. It enables smart routing, same-model fallback, and per-request cost observability without requiring changes to existing SDKs or code. Key Features: Unified API: One base URL and API key for all supported models; compatible with OpenAI SDK, Anthropic SDK, Codex CLI, and Claude Code. Smart Routing & Fallback: Automatically selects eligible provider routes for the requested model and falls back to another provider serving the same model on retryable errors, ensuring continuity. Cost Observability: Detailed traces per request showing final model, provider, token usage, cost, latency, and status; per-API-key cost tracking and spend caps. Flexible Billing: Pay-as-you-go wallet top-ups via credit card, Alipay, or stablecoins (USDT/USDC); optional monthly subscription tiers with shared request quotas. China Access: Direct connectivity from mainland China without VPN, with published latency benchmarks. Data Privacy: Prompts and responses are forwarded in real-time and not persisted; only metadata retained per data retention policy. Target Audience: AI application developers, teams building AI agents, companies using coding assistants (Claude Code, Codex CLI), and organizations needing reliable multi-model LLM access with cost control. Value Proposition: Reduce vendor lock-in, improve resilience via automatic fallback, gain full visibility into LLM spend, and simplify integration with existing tools.