Build AI Apps with Python: Vision — AI That Can See Images | Episode 6

Claude can see! In this episode, we send images to Claude and get back detailed descriptions. Read any image file, convert it to base64, and send it alongside a text prompt — Claude analyzes the image and tells you what it sees. This completes Phase 1 of the series. You now know all the core API capabilities: text, system prompts, conversations, streaming, structured output, and vision. Next up: giving AI the ability to call your functions. Student code: https://github.com/GoCelesteAI/build-ai-apps-python/tree/main/episode06 Every keystroke is shown on screen with 3-second pauses so you can follow along at your own pace. What You'll Learn: • base64 encoding — converting images to text strings for APIs • Multi-modal messages — image block + text block in one message • The image source structure: type, media_type, data • Building a reusable describe_image() function • Handling different image formats (PNG, JPEG) • Sending two different images with the same function • Running Python scripts with :!python % Timestamps: 0:00 - Introduction 0:12 - Claude Can See Images (Preview) 0:44 - Image 1: Sunset 0:49 - Image 2: City Skyline 0:54 - Creating vision.py 1:10 - Imports (new: base64 module) 1:50 - Setup 2:10 - describe_image() function 2:28 - Reading image files in binary mode 2:48 - base64.b64encode — the key conversion 3:05 - Media type detection (PNG vs JPEG) 3:35 - Save progress 3:55 - API call with image block 4:38 - Image source: base64, media_type, data 5:48 - Text block: "Describe this image in detail" 6:22 - Close brackets and return 7:00 - First call: sunset.png 7:28 - Second call: city.png 7:52 - Save and run 8:18 - Claude describes both images! 9:40 - Code review 10:00 - Recap: 3 Key Takeaways 10:32 - End Screen Key Takeaways: 1. base64 encoding converts images to text strings the API can handle 2. Message content becomes a list — image block plus text block together 3. Claude sees and understands images — same API, new capability This completes Phase 1 (API Fundamentals) of Build AI Apps with Python in Neovim. Next: Phase 2 — Tool Use & Function Calling (Episodes 7-11).