Some questions are easier to show than to type. What is this plant, what does this error message mean, how do I read this form. All four of these tools can look at a picture and answer, and each is good at a slightly different kind of “look at this.” Here is one image feature per tool, with the official steps and an honest note on when it helps.
Claude: read the text and detail inside a picture

Claude can look at a photo you upload and reason about what is in it, and it is especially good at the words inside an image.
- Open a new chat at claude.ai or in the Claude app.
- Click the plus to add a photo, or just drag one into the message box. On your phone you can take a picture instead.
- Ask your question in plain words, for example “what does this letter say I owe, and by when.”
- Keep going in the same chat. You can add more images if you need to (up to 20). Claude reads JPEG, PNG, GIF, and WebP.
When to reach for it: when the answer is buried in the image itself, like messy handwriting, a confusing bill or form, a screenshot of an error, or a chart you want explained in plain language.
Source: Claude Help Center
ChatGPT: identify something and get the next step

ChatGPT can look at a photo and help you name what you are seeing and decide what to do about it.
- Open ChatGPT on your phone or at chatgpt.com.
- Tap the plus icon and choose Add photos & files, or drag or paste an image in. On the desktop app you can also choose Take Photo.
- Before you send, you can mark up the photo to circle the exact part you mean.
- Ask your question, for example “what is this plant and is it safe for a cat.”
When to reach for it: when you want to identify something and know what to do next, like a plant, a product label, a rash, or a math problem you want worked out step by step. Point at the part that matters and it focuses there.
Source: OpenAI Help Center
Copilot: point your camera and talk it through

Copilot Vision uses your phone’s live camera instead of a still photo, so you can aim it at something and ask out loud, hands free.
- Open the Copilot app on your phone, on iOS or Android, and sign in with a free Microsoft account.
- Tap the glasses icon to start a Copilot Vision session.
- Point your camera at what is in front of you and ask your question out loud. Copilot answers with its voice.
- End the session when you are done. Nothing you show it during the session is saved.
When to reach for it: when your hands are busy and the thing is right there, like a knob on a broken appliance, a recipe on the counter, or a plaque on a walk. This is the free consumer Copilot, not the paid Microsoft 365 one.
Source: Microsoft Support
Gemini: snap it in the moment and ask

Gemini lets you take a photo right inside the app and ask about it, which is handy when you are out and about.
- Open the Gemini app on your phone.
- Tap Add files. Choose Photos to pick one from your library, or Camera to take a new one.
- Ask your question about the picture, for example “translate this menu and tell me which dishes have no meat.”
- You can add up to 10 images in a single prompt if you are comparing a few things.
When to reach for it: for quick, on-the-go lookups where you take the picture on the spot, like a menu in another language, a label in a store, or what you could cook from what is in the fridge.
Source: Gemini Apps Help
One thing to keep in mind
A camera makes these tools feel confident, but they can still misread a photo, especially small print, handwriting, and numbers. Check anything that matters, like a dose, a due date, or a dollar amount, against the real document. And think before you share a picture that has personal details in it, like an ID or a full account number.
Want this in your inbox each week? Subscribe to Darice on AI.