# Agumbe AI Gateway

Agumbe AI Gateway gives teams one governed API layer for LLM traffic, with OpenAI-compatible chat and embeddings, model routing, app-scoped guardrails, request visibility, provider key management, and spend controls.

Use the gateway when you want application teams to integrate with AI models through a stable enterprise control plane instead of wiring every app directly to every provider.

## What you can do

* Send chat completion requests through `POST /api/v1/llm/chat/completions`.
* Create embeddings through `POST /api/v1/llm/embeddings`.
* Use model aliases such as `smart-default`, `cheap-fast`, `reasoning`, and `embed-default`.
* Scope policy by app with gateway API keys and `agumbe_guardrails_app_id`.
* Apply guardrails for injection detection, PII handling, secrets handling, denied topics, output checks, model allowlists, token limits, and rate limits.
* Inspect request status, latency, token usage, estimated cost, provider, and upstream model in Agumbe Console.
* Manage provider credentials, budgets, wallet balance, and usage controls centrally.

## Start here

1. Read [What is Agumbe AI Gateway?](/agumbe-docs/overview/what-is-agumbe-ai-gateway.md).
2. Make your first request with the [First request](/agumbe-docs/quickstart/first-request.md) quickstart.
3. Learn where teams operate the system in [Gateway and Console](/agumbe-docs/overview/gateway-and-console.md).
4. Use the [API reference](/agumbe-docs/api-reference/authentication.md) for endpoint-level contracts.
5. Use [Operations](/agumbe-docs/operations/deployment.md) when deploying or running the gateway.

## Base URLs

The hosted API examples use:

```
https://api.agumbe.ai
```

The OpenAI-compatible SDK base URL is:

```
https://api.agumbe.ai/api/v1/llm
```

## Quick example

```bash
curl https://api.agumbe.ai/api/v1/llm/chat/completions \
  -H "Authorization: Bearer $AGUMBE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "smart-default",
    "messages": [
      { "role": "user", "content": "Explain what an AI gateway does in one sentence." }
    ],
    "max_completion_tokens": 120
  }'
```

## Documentation model

These docs are organized for four audiences:

* Developers use Quickstart, Guides, SDKs, and API Reference.
* Administrators use Console and Admin sections for keys, guardrails, budgets, and provider credentials.
* Platform teams use Operations for deployment, configuration, observability, storage, and reliability.
* Security and business reviewers use Overview, Admin, and Reference pages for governance and control-plane behavior.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://agumbe.gitbook.io/agumbe-docs/readme.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.