Skip to content

quick-start.md

Latest commit

 

History

History
164 lines (127 loc) · 4.87 KB

quick-start.md

File metadata and controls

164 lines (127 loc) · 4.87 KB
title Quick Start
slug /getting-started/quick-start
description Deploy the AISIX open source AI gateway with Docker in under 5 minutes. Configure LLM models, create API keys, and make your first OpenAI-compatible request.
keywords
deploy AI gateway
LLM proxy setup
quick start AI gateway
LLM API gateway Docker
open source AI gateway

This guide walks you through starting AISIX with a single command and making your first AI request in under 5 minutes.

Prerequisites

Step 1: Start AISIX

Run the following command to download and start AISIX:

curl -fsSL https://run.api7.ai/aisix/quickstart | sh

This script:

  • Downloads docker-compose.yaml and config.yaml to ~/.aisix/
  • Generates a random Admin Key and writes it into the config
  • Pulls and starts AISIX and etcd via Docker Compose

When the script completes, you will see output like:

[aisix] AISIX is running!

  Proxy API:   http://127.0.0.1:3000
  Admin API:   http://127.0.0.1:3001/aisix/admin
  Admin UI:    http://127.0.0.1:3001/ui
  API Docs:    http://127.0.0.1:3001/openapi
  Admin Key:   <generated-admin-key>

  Export it:    export ADMIN_KEY=<generated-admin-key>

Copy the export line and run it in your terminal — all examples in this guide use the $ADMIN_KEY variable:

export ADMIN_KEY=<your-admin-key>

Step 2: Configure a Model

Tell AISIX which upstream LLM to use. This example uses OpenAI's gpt-4o-mini:

curl -X POST http://127.0.0.1:3001/aisix/admin/models \
  -H "Authorization: Bearer $ADMIN_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "openai-gpt4",
    "model": "openai/gpt-4o-mini",
    "provider_config": {
      "api_key": "<YOUR_OPENAI_API_KEY>"
    }
  }'
  • Replace <YOUR_OPENAI_API_KEY> with your OpenAI API key.
  • name is the identifier used in requests to AISIX.
  • model specifies the provider (openai) and the upstream model ID.

A 201 response confirms the model was created.

Step 3: Configure an API Key

Create an API key to authenticate requests to the proxy:

curl -X POST http://127.0.0.1:3001/aisix/admin/apikeys \
  -H "Authorization: Bearer $ADMIN_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "key": "my-secret-key",
    "allowed_models": ["openai-gpt4"]
  }'
  • key is the secret used in the Authorization header when calling the proxy.
  • allowed_models controls which models this key can access.

Step 4: Make Your First Request

Send a chat completion request through AISIX to the upstream provider:

curl -X POST http://127.0.0.1:3000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer my-secret-key" \
  -d '{
    "model": "openai-gpt4",
    "messages": [
      {
        "role": "user",
        "content": "Tell me a fun fact about AI"
      }
    ]
  }'

You will receive a standard OpenAI-compatible chat completion response from the upstream provider, proxied through AISIX.

Example: Other Providers

Adding other providers follows the same pattern — use the correct model prefix and provide the right credentials in provider_config.

Google Gemini

curl -X POST http://127.0.0.1:3001/aisix/admin/models \
  -H "Authorization: Bearer $ADMIN_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-gemini",
    "model": "gemini/gemini-2.5-flash",
    "provider_config": {
      "api_key": "<YOUR_GEMINI_API_KEY>"
    }
  }'

Anthropic

curl -X POST http://127.0.0.1:3001/aisix/admin/models \
  -H "Authorization: Bearer $ADMIN_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-anthropic",
    "model": "anthropic/claude-3-5-sonnet-20241022",
    "provider_config": {
      "api_key": "<YOUR_ANTHROPIC_API_KEY>"
    }
  }'

DeepSeek

curl -X POST http://127.0.0.1:3001/aisix/admin/models \
  -H "Authorization: Bearer $ADMIN_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "my-deepseek",
    "model": "deepseek/deepseek-chat",
    "provider_config": {
      "api_key": "<YOUR_DEEPSEEK_API_KEY>"
    }
  }'

After creating a new model, update your API key's allowed_models list to grant access to it.

Next Steps

  • Manage Models via UI: AISIX includes a built-in dashboard for managing LLM models and API keys without curl. See Admin UI.
  • Add More Providers: Add Anthropic, Gemini, or DeepSeek models using the same pattern — see Model Management for a full CRUD reference.
  • Control LLM Costs: Configure RPM, TPM, and concurrency limits to prevent runaway usage. See Rate Limiting.
  • Secure Access: Understand how API key authentication and model allowlists work in Authentication.