OpenAI¶

The OpenAI provider for AiKit allows you to use GPT models like GPT-4o and GPT-4o mini.

Setup¶

By default, the OpenAI provider uses the OPENAI_API_KEY environment variable.

export OPENAI_API_KEY="your-api-key-here"

Language Models¶

You can resolve OpenAI models using the llm() registry function.

import { llm, stream, Message } from "@codeworksh/aikit";

// Initialize the model
const model = await llm("openai", "gpt-4o");

Example Usage¶

const context: Message.Context = {
    messages: [
        Message.createUserMessage({
            role: "user",
            time: { created: Date.now() },
            parts: [{ type: "text", text: "Explain recursion briefly." }],
        }),
    ],
};

const message = await stream.complete(model, context);
console.log(message.parts);

Provider Options¶

For fine-grained control, you can pass provider-specific options such as reasoning effort for models like o1 or o3-mini.

await stream.complete(model, context, {
    apiKey: "custom-openai-key",
    // Example: Set reasoning effort
    reasoningEffort: "medium",
});