Provider Clouds
Every provider cloud behind one nRouter key: learn what runs on OpenAI, Anthropic, AWS Bedrock, Google Vertex AI, Azure AI Foundry, and Alibaba US.
nRouter unifies major AI provider clouds behind a single API endpoint and virtual key. Models across OpenAI, Anthropic, AWS Bedrock, Google Vertex AI, Azure Foundry, and Alibaba US sit in a shared catalog, accept identical request shapes, settle at exact list price with 0% platform fee, and share uniform budget, guardrail, and telemetry controls.
| Provider cloud | Example models | Known for |
|---|---|---|
| OpenAI | gpt-5.5, gpt-4o, o1 | Text generation, structured outputs, vision, and embeddings |
| Anthropic | claude-sonnet-4-5-20250929, claude-3-5-haiku | Reasoning, code analysis, long context, and native Messages |
| AWS Bedrock | Llama 3.3, Amazon Nova, Mistral Large 3 | Multi-cloud resilience without separate AWS credentials |
| Google Vertex AI | gemini-2.5-pro, gemini-2.5-flash | Multimodal comprehension, audio, and 2M token context windows |
| Azure Foundry | GPT-4o, Phi-4, Mistral Large | Enterprise compliance posture and direct cloud egress |
| Alibaba US | Qwen 2.5, DeepSeek R1, DeepSeek V3 | High-efficiency open-weight reasoning and code generation |
Uniform Platform Guarantees
One sk-nrouter-* virtual key calls every cloud via standard Chat Completions; no provider accounts needed. Every model settles at its provider list price with 0% platform markup. Budgets, guardrails, and audit logging apply identically across clouds. See Analytics.
Frequently asked questions
Do I need an account with each cloud?
No; your nRouter key is the only credential needed, and nRouter does not require BYOK.
Does the cloud a model runs on change its price?
No; all models settle at exact published provider list rates with 0% markup.
How do I see one cloud's usage on its own?
Filter by provider dimension in dashboard Analytics or inspect request logs.