Back to Workbench

AI Sound Identifier

Building an AI based animal sound identifier

Animal sound identifier powered by Gemini 3. Image generated with Nano Banana.

This is a quick POC that originated as a fun game I was playing with my son - one makes an animal sound, and the other guesses. We decided to test Gemini on the phone, but realized it relies on Speech-to-Text conversion and cannot accept raw audio inputs. I had also learnt about Gemini CLI, and wanted to ditch the IDE and try building something natively in the CLI.