calories

OpenAI ChatGPT helper function This val uses your OpenAI token if you have one, and the @std/openai if not, so it provides limited OpenAI usage for free. import { chat } from "https://esm.town/v/stevekrouse/openai"; const { content } = await chat("Hello, GPT!"); console.log(content); import { chat } from "https://esm.town/v/stevekrouse/openai"; const { content } = await chat( [ { role: "system", content: "You are Alan Kay" }, { role: "user", content: "What is the real computer revolution?"} ], { max_tokens: 50, model: "gpt-4o" } ); console.log(content);

An interactive, runnable TypeScript val by stevekrouse

File to Data URL Helpers to convert files to base64 & base64-encoded data URLs, which are particularly helpful for sending images to LLMs like ChatGPT, Anthropic, and Google. ChatGPT Live example import { fileToDataURL } from "https://esm.town/v/stevekrouse/fileToDataURL"; const dataURL = await fileToDataURL(file); const response = await chat([ { role: "system", content: `You are an nutritionist. Estimate the calories. We only need a VERY ROUGH estimate. Respond ONLY in a JSON array with values conforming to: {ingredient: string, calories: number} `, }, { role: "user", content: [{ type: "image_url", image_url: { url: dataURL, }, }], }, ], { model: "gpt-4o", max_tokens: 200, }); Anthropic Live example import { fileToBase64 } from "https://esm.town/v/stevekrouse/fileToDataURL"; const base64File = await fileToBase64(file); let res = await anthropic.messages.create({ model: "claude-3-5-sonnet-20240620", max_tokens: 1024, messages: [ { "role": "user", "content": [ { "type": "image", "source": { "type": "base64", "media_type": file.type, "data": base64File, }, }, { "type": "text", "text": `Write an HTML email that evokes this photo in the funniest way possible, with code fences.`, }, ], }, ], }); Google Live example import { fileToBase64 } from "https://esm.town/v/stevekrouse/fileToDataURL"; const base64Image = await fileToBase64(image); const result = await model.generateContent([ "Write all the names and authors of these books in JSON format. The response should be a valid JSON array of objects, each with 'title' and 'author' properties.", { inlineData: { data: base64Image, mimeType: image.type, }, }, ]);

Code from https://deno.com/blog/build-image-resizing-api Useful for compressing an image before sending to chatgpt4v, for example