> ## Documentation Index
> Fetch the complete documentation index at: https://www.edgee.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# LLM Router and SDKs

> Connect an application to Edgee using an existing client, an Edgee SDK, or HTTP.

Use Edgee as an **LLM router** when your application sends model requests directly through an API client, an SDK, or HTTP. Configure a Gateway API key and a client, then inspect the resulting requests in the console.

If you want to run Claude Code, Codex, Cursor, or another coding assistant through Edgee, start with the [coding assistant quickstart](https://www.edgee.ai/docs/quickstart/index.md).

## Choose your integration

| Your starting point | Guide |
| - | - |
| Existing OpenAI client | [OpenAI SDK](https://www.edgee.ai/docs/llm-router/integrations/openai.md) |
| Existing Anthropic client | [Anthropic SDK](https://www.edgee.ai/docs/llm-router/integrations/anthropic.md) |
| LangChain application | [LangChain](https://www.edgee.ai/docs/llm-router/integrations/langchain.md) |
| OpenClaw configuration | [OpenClaw](https://www.edgee.ai/docs/llm-router/integrations/openclaw.md) |
| New application using an Edgee SDK | [TypeScript](https://www.edgee.ai/docs/llm-router/sdk/typescript.md), [Python](https://www.edgee.ai/docs/llm-router/sdk/python.md), [Go](https://www.edgee.ai/docs/llm-router/sdk/go.md), or [Rust](https://www.edgee.ai/docs/llm-router/sdk/rust.md) |
| Direct HTTP requests | [First API request](https://www.edgee.ai/docs/llm-router/quickstart/api-key.md) and [Gateway API reference](https://www.edgee.ai/docs/llm-router/api-reference.md) |

## Send your first request

1. [Sign in to the console](https://www.edgee.ai/login), or accept your team’s invitation.
2. [Create a Gateway API key](https://www.edgee.ai/docs/llm-router/quickstart/api-key.md) and store it in your application's secret environment.
3. Choose a client above and configure the appropriate base URL. OpenAI clients usually use `https://edgee.io/v1`; follow the integration guide for your client.
4. Send a model request, then open **Logs** in the same organization to verify the key, served model, provider, and usage.

For the complete sequence, follow the [application quickstart](https://www.edgee.ai/docs/llm-router/quickstart.md).

## Configure application usage

| Task | Guide |
| - | - |
| Restrict or rotate application credentials | [Gateway API keys](https://www.edgee.ai/docs/llm-router/api-keys.md) |
| Track tokens, costs, and request tags in code | [Application observability](https://www.edgee.ai/docs/llm-router/observability.md) |
| Use your own provider account | [BYOK](https://www.edgee.ai/docs/features/byok.md) |
| Allow models and provider routes | [Model access](https://www.edgee.ai/docs/features/model-access.md) |
| Configure routing and backups | [Retry and fallback](https://www.edgee.ai/docs/features/retry-and-fallback.md) |
| Cap application spending | [Usage limits](https://www.edgee.ai/docs/features/usage-limits.md) |

Organization settings also apply to application traffic. Review inherited settings before changing an individual key.

## Gateway API versus Console API

The **Gateway API** serves model requests and uses Gateway API keys. The **Console API** manages organization resources and uses personal API tokens. To automate key management or export usage, use the separate [Console API](https://www.edgee.ai/docs/api-reference/console-api-setup.md) section.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.