32dots HEIDELBERG AI
Session 4 medium

Built-in AI: text, images, speech & file extraction

LESSONLesson 4 · ~20 min

🎯Goal. Add real AI features — text, images, speech and document data-extraction — with no API keys and no model accounts.

▶ Try this prompt

Add a button that uses AI to suggest the best free room for a given time and group size. And when a booking is confirmed, email the person a confirmation.

Base44 ships AI as built-in functions — invokeLLM (text), GenerateImage, GenerateSpeech, SendEmail, and ExtractDataFromUploadedFile. No keys.

  1. 1Send the prompt above. Base44 wires the built-in invokeLLM (an AI call) and SendEmail directly — no key paste, no model account.
  2. 2Other built-ins you can ask for: GenerateImage, GenerateSpeech (text → audio), and ExtractDataFromUploadedFile (turn an uploaded PDF/CSV/image into structured rows).
  3. 3Watch your integration credits: roughly — an LLM call ~3, an email ~1, an image ~1, speech ~1 per 50 characters.

You'll see. An AI suggestion and a confirmation email working inside your app — built in, not bolted on with keys.

💳Cost. Built-in AI/email/image features consume integration credits (free plan: 100/month). No external API keys or accounts.

💡Takeaway. Base44 adds intelligence by description, with no keys to manage — but each AI call quietly spends integration credits.