The AI Gateway is an advanced system designed to facilitate communication with over 200 models in language processing, vision, audio, and more, all through a single, efficient API. Whether you’re looking to quickly set up a hosted solution or self-host the open-source version, this guide will walk you through the steps, ensuring you get the most out of your AI Gateway experience.

Getting Started with AI Gateway

There are two primary options for using the AI Gateway: utilizing the hosted API on portkey.ai, or self-hosting the open-source version. Below are the steps for each method:

1. Hosted Gateway on portkey.ai

The fastest way to implement an AI Gateway for your generative AI application is via the hosted API. Here’s how:

• Visit the Portkey AI homepage.
• Sign up and acquire your API key.

2. Self-hosting the Open Source Version

If you prefer to run the AI Gateway locally, follow these instructions:

• Make sure you have npx installed.
• Run this command in your terminal:

npx @portkey-ai/gateway sup

• Your AI Gateway will now be operational at http://localhost:8787.

3. Self-hosting the Enterprise Version

The enterprise version offers additional benefits such as enhanced security and governance. For comparisons between the open-source, hosted, and enterprise versions, click here.

Making Requests Through the AI Gateway

Once your Gateway setup is complete, you can make requests in various ways:

• OpenAI API SDK: Incorporate the Gateway URL into your OpenAI SDK and make requests as usual. The Gateway will translate requests to fit various provider formats.
• Python SDK: Utilize the Portkey Python SDK by running:

pip install -qU portkey-ai

• Node.JS SDK: Use the Portkey JS SDK by running:

npm install --save portkey-ai

• REST APIs: The AI Gateway supports OpenAI compatible endpoints with extra parameters for broader accessibility.

Advanced Features of AI Gateway

The AI Gateway stands out due to its advanced features, which include:

• Load Balancing: Distributes LLM requests across multiple API keys or providers for optimal performance.
• Automatic Retries: Automatically retries failed requests, with up to five attempts using an exponential backoff strategy.
• Request Timeouts: Allows you to manage latencies by setting granular request timeouts, terminating requests that exceed specified durations.
• Multimodal Capabilities: Supports various model types, including vision and audio, in addition to text models.

Troubleshooting

If you encounter any issues while setting up or using the AI Gateway, consider the following troubleshooting tips:

• Ensure all dependencies and libraries are properly installed.
• Verify your API keys are correct and active.
• Check network connectivity and firewall settings that may block requests.
• Consult the [documentation](https://docs.portkey.ai/docs) for specific error codes and their resolutions.

For more insights, updates, or to collaborate on AI development projects, stay connected with [fxis.ai](https://fxis.ai/edu).

Conclusion

Setting up the AI Gateway opens a world of possibilities for leveraging advanced AI models seamlessly. With its robust feature set, whether you’re aiming for a quick hosted solution or a fully customized enterprise setup, you can empower your applications to perform at their best. At [fxis.ai](https://fxis.ai/edu), we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.