Welcome to WRouter

WRouter is an OpenAI-compatible LLM gateway that lets you call almost every major model through a single key and a single base URL.

Core capabilities

• Unified protocol — Mainstream LLM formats auto-translated; flip one line of base_url in any official SDK and you're in.
• Multi-vendor aggregation — Dozens of upstream providers worldwide and in China — pay-as-you-go, switch freely, no per-vendor signup.
• Full multi-modal — Chat, embeddings, images, TTS / STT, video — one auth, one bill.
• Enterprise controls — Sub-tokens, per-token quotas, IP allowlists, call logs, spend dashboards, rate-limit alerts.
• Developer-friendly — Zero migration cost, drop-in with the official SDKs, full bilingual docs with code examples.

Why WRouter

Use case	Value
Multi-model orchestration	One codebase, one key reaches every major provider — no extra accounts, SDKs, or invoices
Cross-vendor benchmarking	Same API surface; swap the upstream to A/B test and regress with zero migration
Team collaboration	Sub-tokens with quotas, IP allowlists, call logs, spend dashboards — clear and auditable
Coding-agent integration	Configure once; every popular IDE and CLI agent works — no per-tool key setup
Multi-modal in one place	Chat, embeddings, images, TTS / STT, and video share auth and billing

Next steps

• Quick Start — Sign-up to first call in 5 minutes
• Billing — Top-up, pay-as-you-go, plans
• API Reference

Resources

• Console: https://wrouter.ai/dashboard/overview
• API base URL: https://wrouter.ai
• Model marketplace: https://wrouter.ai/models
• GitHub Issues: https://github.com/wrouter/wrouter/issues