What it does
AI Vision explains a selected screenshot, extracts text from images, summarizes supported webpages, and compares supported tabs. Follow-ups stay with the answer.
See a screenshot example →AI SCREENSHOT ASSISTANT FOR CHROME · PUBLIC STORE V2.5
Your AI screenshot assistant for Chrome.
Explain screenshots, copy text from images, and ask follow-up questions in Chrome.
Free extension · Your Gemini API key required · Get your Gemini API key
PRODUCT FACTS
What AI Vision does in Chrome.
AI Vision is a free Gemini Chrome extension for asking about a screenshot, image, supported webpage, or supported tab with your own API key.
Best for desktop Chrome screenshots, image-to-text, webpage summaries, and tab comparisons on supported webpages. Chrome internal pages and the Web Store are restricted.
A referral does not connect an assistant account. Whether you arrive from ChatGPT, Gemini, Claude, Kimi, DeepSeek, Grok, Z.ai, Perplexity, or Copilot, AI Vision still uses Gemini with the key you provide in Chrome.
Who it suits
Use it for charts, error messages, image text, articles, product research, and other visual questions while browsing.
See text-extraction examples →Where it works
Use the free extension in desktop Google Chrome on supported webpages. Chrome internal pages and the Web Store cannot be analyzed.
Read the webpage guide →Your key, your limits
Bring a Gemini API key from Google AI Studio. Google sets model access, quotas, and any billing separately from installation.
Get a Gemini API key →Free to install
Install AI Vision from the official Chrome Web Store. The extension is free; Gemini usage follows Google’s project limits.
Open the official listing ↗Privacy and source
Your key and preferences stay in Chrome storage. Requests go directly to Google’s Gemini API; review the privacy notice and public GitHub repository.
Read the privacy notice →View the source ↗TRY YOUR FIRST SCREENSHOT HERE
Select. Explain. Follow up.
Choose a sample to see the same idea as the installed extension. Each answer is fictional; the demo never sends an API request.
Fictional screenshot · Reading club
Pages read each week
406090100
One-tap explanation
“Explain this screenshot.”
Illustrative answer
Reading increased most from week 2 to week 3: from 60 to 90 pages, a gain of 30. Week 4 has the highest total, but the increase is smaller at 10 pages.
Follow up: “What would I need to know before calling this a trend?”
Learn to explain a screenshot →Fictional article · Community library
A quieter space after school
The library is testing a quiet study hour from 4–5 pm on Tuesdays and Thursdays for six weeks. The ground floor stays open for conversation. Staff will collect visitor feedback before deciding whether to continue.
You ask
“What’s changing, and what’s still undecided?”
Illustrative answer
Quiet study hours are being trialed twice a week for six weeks. The ground floor remains a social space. Whether the quiet hours become permanent depends on visitor feedback.
Learn to summarize a webpage →Fictional product tabs
Two bags. One decision.
Daylight 18: 18 L capacity, 600 g, padded laptop sleeve, no rain cover.
Cloudpack 22: 22 L capacity, 850 g, rain cover included, laptop sleeve not specified.
You ask
“Which bag fits a light laptop commute?”
Illustrative answer
Daylight 18 is 250 g lighter and explicitly includes a laptop sleeve. Cloudpack 22 adds 4 L of space and rain protection, but you’d need to verify laptop storage before choosing it.
Learn to compare Chrome tabs →Fictional content and illustrative responses. No API call, key, or page access is used by this demo.
Try AI Vision free ↗Gemini API key required · Get set up first
Three steps to your first answer.
Install AI Vision, save your Gemini key, then return to the sample above and try a normal webpage.
Try your first screenshot here →Get your Gemini API key →Add AI Vision to Chrome
Install the public v2.5 build from the Chrome Web Store. AI Vision cannot run on the Web Store itself.
Get and save your Gemini key
In Google AI Studio, create a key, paste it into AI Vision Settings, and save it. Google controls Gemini limits and pricing separately from installation.
Open Google AI Studio ↗Return, capture, and ask
Open a normal webpage, click AI Vision, choose the public Capture mode, drag around the detail that matters, and type a question or choose Explain. Follow up from the answer.
Choose the context you need.
Use a screenshot for a visual detail, a page for its key points, or tabs for a side-by-side comparison.
Explain or extract
Understand charts, diagrams, errors, and text inside images.
Screenshot guide →Copy text guide →Summarize a page
Turn a readable article into a summary you can check.
Webpage guide →Compare tabs
Bring related sources together and keep their differences visible.
Tab comparison guide →Simple to use.
Clear about the details.
Your key stays in Chrome. Requests go directly to Google. Optional browser tasks keep their approval and Stop controls.
Read the privacy notice →Do I need a Gemini API key?
Yes. AI Vision uses your own key from Google AI Studio. Save it in the extension’s Settings. The website never asks you to paste a key.
Is AI Vision free?
The extension is free to install. Google controls Gemini API pricing, model availability, and usage limits. Check the limits and billing for your project in Google AI Studio.
What gets sent to Google?
Your question and the screenshot or page context needed for the request go directly to the Gemini API. Your key stays in Chrome’s local extension storage and is sent to Google to authenticate requests. There is no developer proxy or analytics.
Does it work on every page?
AI Vision works on supported webpages. Chrome internal pages, the Chrome Web Store, and other restricted pages cannot be analyzed. Compare tabs needs optional permission and reads up to 20 supported tabs in the current window.
Start with the screenshot in front of you.
Install AI Vision, save your Gemini key, and get a clear answer with a follow-up ready.