Assistant API Call without using Threads

I want to use an assistant that I’ve created but I want to handle all the message and thread management myself.

Is there an API call I can make that similar to a vanilla open ai api call but having the response come from my assistant?

I’m looking for something like this but also being able to specify the assistant ID.

const apiRequestBody = {
      model: "gpt-3.5-turbo",
      messages: [{ role: "system", content: systemPrompt }, ...apiMessages],
    };

    const response = await fetch(openAIUrl, {
      method: "POST",
      headers: {
        "Authorization": `Bearer ${openAI_API_Key}`,
        "Content-Type": "application/json",
      },
      body: JSON.stringify(apiRequestBody),
    });
1 Like