Hamel Husain built a free plugin that connects Claude Code to Gemini, letting Claude analyze images, videos, PDFs, and audio by requesting Gemini's help. Setup takes two minutes: get a free API key from ai.dev and install with two commands. Claude automatically calls Gemini when it needs to process multimodal files.
Claude is great at writing code. It's not great at understanding images, videos, and PDFs. Here's how to give it a model that is. Gemini is the best multimodal model. I built a plugin that connects it to Claude code. Need references to a paper, a video, an audio recording, or all of the above at once? Now your agent can see it. Claude can send Gemini off in the background to study a long video or document. It reports back just what's relevant so it doesn't bloat your context window. Analyzing YouTube videos are especially fast fast because Gemini has a special integration that just works. Setting it up takes 2 minutes. Go to ai.dev, sign in with Google, and grab a free API key. Gemini's free tier is very generous. Install my free plugin with just 2 commands and Claude learns when to call Gemini on its own. Ask it to look at a file and it handles the rest. Comment gem and I'll send you the setup instructions.
