Claude Clue #22: Feed Claude More Than Text

Claude Code can already see images and PDFs—you paste them in with Alt+V on Windows or Ctrl+V on Mac. But it doesn't stop there. I recently ran a discovery session where we collected hours of audio recordings, photos of whiteboards, scanned documents, and messy handwritten notes. The kind of sprawling, multi-format chaos that normally lives in a folder you never open again. Claude Code handled all of it, once I gave it the right tools.

For audio and video, just ask Claude if it has the libraries to process the file. If not, you can install something like openai-whisper (for audio) and Claude will use it to transcribe recordings, summarize them, or even clean up transcription errors. The result is a more complete picture—Claude working with the same raw material you collected, not just the parts you had time to type up. If you want to know what else it can process, ask it. Claude's surprisingly good at explaining its own capabilities.

AI-assisted.

HT: Tim "Roscoe" Rawson