Each LLM is an adapter around a language model provider and the specific model version, eg: gpt-4o-mini. Each Agent can pick their own model and a ZeeWorkflow can be configured to use a specific LLM as default.

const llm = new LLM({
    provider: "openai",
    id: "gpt-4o-mini",
});

​
List of supported LLM Model IDs

"gpt-4o"
"gpt-4o-mini"
"gpt-4o-2024-05-13"
"gpt-4o-2024-08-06"
"gpt-4o-2024-11-20"
"gpt-4o-audio-preview"
"gpt-4o-audio-preview-2024-10-01"
"gpt-4o-audio-preview-2024-12-17"
"gpt-4o-mini-2024-07-18"
"gpt-4-turbo"
"gpt-4-turbo-2024-04-09"
"gpt-4-turbo-preview"
"gpt-4-0125-preview"
"gpt-4-1106-preview"
"gpt-4"
"gpt-4-0613"
"gpt-3.5-turbo-0125"
"gpt-3.5-turbo"
"gpt-3.5-turbo-1106"

​
Environment Variables

OPENAI_API_KEY

​
Use Cases

​
Image Analysis

LLMs can also process images along with text using image URL messages. Here’s an example of analyzing an image using the LLM:

const messages = [
    userMessage(
        "What's in this image?"
    ),
    userMessage([
        {
            type: "image",
            image: "https://example.com/logo.png",
        },
    ]),
],

const schema = z.object({
    description: z.string(),
    colors: z.array(z.string()),
    text_content: z.string().optional(),
    improvements: z.string().optional(),
});

const result = await llm.generate({
    messages,
    schema,
    temperature: 0.8,
});

The LLM will analyze the image and return a structured response containing the description, colors used, and potential improvements. You can also use base64-encoded images by providing the data URL:

const messages = [
    userMessage("What's in this image?"),
    userMessage([
        {
            type: "image",
            image: "data:image/png;base64,...",
        },
    ]),
];