Inter-office memo

To
Everyone who wants their coding agent to see their screen
From
Mo, Chief Executive
Date
November 18, 2026, 10:43 AM CST
Re
Screenshots in Claude Code and Codex

How to paste a screenshot into Claude Code or Codex in 2026: 8 ways compared

In Claude Code’s terminal, press Ctrl+V, even on a Mac outside iTerm2, or Alt+V on Windows and WSL; in the Codex CLI, paste or run codex -i screenshot.png. When paste fails, drag the file in or type its path. Claude’s and ChatGPT’s desktop apps attach images with a button or a drag, Remote Control takes photos from your phone, and Enjoy takes them for every lab’s agent.

Disclosure: Enjoy is our product. We checked every way below against its maker’s own pages on November 18, 2026, and we say where another way is the better pick, including the free ones that don’t need Enjoy.

The ways to give Claude Code or Codex a screenshot, compared

A screenshot of the error, the broken layout or the design you want is often the fastest way to explain a problem to a coding agent, and both Claude Code and Codex read images. The catch is the terminal. It pastes text, so each agent takes images with a key of its own, and that key changes with your system: Anthropic’s shortcut table lists Ctrl+V, Cmd+V only in iTerm2, and Alt+V on Windows and WSL. When a paste key fails, a file path still works. The apps with a window of their own take images with a button or a drag.

Way Best for Price, checked November 18, 2026 Runs on How you add the image
Paste in Claude Code’s CLI Claude Code in a terminal Included in Claude Pro ($20 a month), Max ($100), Ultra ($200), Team and Enterprise Mac, Windows, Linux Ctrl+V; Cmd+V in iTerm2; Alt+V on Windows and WSL
Paste in the Codex CLI, or codex -i Codex in a terminal or a script Included in ChatGPT Plus ($20 a month) and up, or pay as you go with an API key Mac, Windows, Linux Paste into the composer, or pass the file with -i
A file path, typed or dragged Any agent, any terminal, WSL and SSH Free Wherever the agent runs Drag the file in, or type its path
The Claude desktop app Claude Code without a terminal Included in Claude’s paid plans, from $20 a month Mac, Windows; Linux in beta The + button, or drag and drop
The ChatGPT desktop app, for Codex Codex without a terminal Every ChatGPT plan, from Free; Plus is $20 a month Mac, Windows; Linux in preview Hold Shift and drag
The VS Code extensions Either agent beside your code Claude Code’s: a paid Claude plan; Codex’s: Plus and up VS Code and editors built on it, such as Cursor Paste, for Claude Code; Shift and drag, for Codex
Remote Control, in the Claude app A photo from your phone, for Claude Code on your computer Claude Pro and up; not with an API key iPhone, Android, any browser Attach a photo
Enjoy Every lab’s agent, from your computer, a browser or your iPhone A monthly subscription, free for 7 days (pricing) Mac, Windows, Linux; the web; iPhone Paste, drag or +; Take photo on iPhone

Prices are per month as each company lists them, before tax. Claude Code needs a paid Claude plan, and Claude’s pricing has Pro at $20 a month, or $17 billed yearly. Codex’s pricing lists Free at $0, Go at $8, Plus at $20 and Pro from $100, and Plus is where “Codex on the web, in the CLI, in the IDE extension, and on iOS” begins.

First, copy the screenshot instead of saving it

The paste keys read your clipboard, so the screenshot has to be there rather than in a file on your desktop. On a Mac, Apple says to “press Control in addition to the other keys you press to take the screenshot”: Shift-Command-4 with Control copies the part of the screen you select. On Windows, Print Screen takes “a static image snapshot of your entire screen and place[s] it in your clipboard,” and Windows logo key + Shift + S opens the Snipping Tool overlay for part of it. If the screenshot is already a file, skip the clipboard and give the agent its path.

Paste in Claude Code’s CLI

Best for: Claude Code in a terminal.

Anthropic’s common workflows list three ways to add an image: “Drag and drop an image into the Claude Code window,” “Copy an image and paste it into the CLI with Ctrl+V, or with Alt+V on Windows and WSL,” and “Provide an image path to Claude.” Pasting “Inserts an [Image #N] chip at the cursor so you can reference it positionally in your prompt,” so you can write “the button in [Image #1] overlaps the one in [Image #2].” “You can work with multiple images in a conversation,” and “When Claude references images (for example, [Image #1]), Cmd+Click (Mac) or Ctrl+Click (Windows/Linux) the link to open the image in your default viewer.”

On a Mac, the key is the surprise: the shortcut table lists Cmd+V for iTerm2 only, so everywhere else, Apple’s Terminal included, press Ctrl+V.

Price: “Claude Code is included in all paid plans,” says Claude’s pricing: Pro at $20 a month, or $17 billed yearly, Max at $100 and Ultra at $200, plus Team and Enterprise. The Free plan doesn’t include it.

Watch for: on Windows the key is Alt+V, and “On WSL, both Ctrl+V and Alt+V are bound; use Alt+V if your terminal intercepts Ctrl+V.” The CLI takes images, not other attachments: Anthropic’s comparison of the CLI and the desktop app lists file attachments as “Not available” in the CLI, against “Images, PDFs” in the app, so give a PDF by its path.

Paste in the Codex CLI, or codex -i

Best for: Codex in a terminal, and scripts that send a screenshot with the prompt.

OpenAI’s image inputs page says to “Paste an image into the interactive composer, or pass one or more files on the command line”:

codex -i screenshot.png "Explain this error and suggest the smallest fix"
codex --image before.png,after.png "Compare these states and list the regressions"

“For multiple images, separate paths with commas or repeat --image.” OpenAI’s docs don’t name the paste key. Codex’s source code takes Ctrl+V or Alt+V, and Shift+Insert on Windows. In WSL, where the Windows clipboard is often out of reach, its clipboard code has “a WSL fallback that calls PowerShell on the Windows side to write the clipboard image to a temporary file.”

Price: the CLI starts at ChatGPT Plus, $20 a month, and Pro has “Plans at $100, $200, or $500 USD per month.” Free and Go get Codex only as “GPT-6 Luna at Standard speed in the desktop app, subject to rollout.” With an API key, you “Pay for Codex usage based on API pricing.”

Watch for: the paste keys come from the source, not the docs, so a release can change them; -i is the documented way. Codex “accepts common image formats, including PNG and JPEG.”

A file path, typed or dragged

Best for: any agent in any terminal, and the times a paste key won’t work, as in WSL or over SSH.

Every agent can read a picture from disk. Anthropic’s example is “Analyze this image: /path/to/your/image.png,” OpenAI’s image inputs page says you can “ask ChatGPT to inspect an image on your system,” and the Codex CLI’s -i takes a path. You don’t have to type it: on a Mac, Apple says “Drag the file or folder to the Terminal window. The item’s absolute path appears on the command line.”

Price: free.

Watch for: the path has to be one the agent’s computer can open. In WSL, a screenshot saved on Windows is under /mnt/c: Microsoft’s example is that C:\Users\<user name>\Project “will look like this when mounted in a WSL command line: /mnt/c/Users/<user name>/Project.” Over SSH, the agent runs on the other machine, so copy the file there first, with scp for instance, and give that path.

The Claude desktop app

Best for: Claude Code without a terminal, and PDFs as well as images.

Claude Code is the Code tab of Anthropic’s desktop app, and its docs say you can “attach images, PDFs, and other files to your prompt from the + button next to the prompt box, or drag and drop files directly into the prompt. This is useful for sharing screenshots of bugs, design mockups, or reference documents.” No key to remember, and nothing to set up in your terminal.

Price: included in Claude’s paid plans, from Pro at $20 a month.

Watch for: it runs only Claude. “Linux support for the Claude desktop app is in beta,” for Ubuntu 22.04 or later and Debian 12 or later. In a WSL session, “if the + menu shows Add image instead of Add files or photos, you can attach only images.”

The ChatGPT desktop app, for Codex

Best for: Codex without a terminal.

The ChatGPT desktop app is where Codex runs on your computer. To give it a screenshot, OpenAI’s image inputs page says to “Drag an image into the prompt composer while holding Shift to include it as context,” and you can also have it “use a screenshot tool to verify work in another app.” In a prompt, OpenAI’s advice is to “Name what the image shows, point to the area that matters, and state the output and constraints.”

Price: every ChatGPT plan from Free includes it. Free and Go get “GPT-6 Luna at Standard speed in the desktop app, subject to rollout”; Plus, at $20 a month, adds GPT-6.1 Sol.

Watch for: it runs only Codex, and the drag needs Shift. It’s for macOS and Windows, and “The ChatGPT desktop app for Linux is available in preview.”

The VS Code extensions

Best for: Claude Code or Codex beside your code, in VS Code or an editor built on it, such as Cursor.

Anthropic’s extension takes a paste: “To attach an image, paste it from your clipboard into the prompt box,” and “To attach files, hold Shift while dragging them into the prompt box.” It needs VS Code 1.94.0 or later. OpenAI’s Codex extension takes a drag: “Drag an image into the prompt composer while holding Shift so the extension accepts the drop instead of passing it to the editor.”

Price: Claude’s extension comes with a paid Claude plan, from $20 a month, and Codex’s with ChatGPT Plus and up, also from $20.

Watch for: each extension runs one lab’s agent, and OpenAI’s docs describe only the drag for Codex’s, not a paste.

Remote Control, in the Claude app

Best for: a photo or screenshot from your phone, for Claude Code running on your computer.

Start a session with claude remote-control in your project, or /remote-control in the desktop app or VS Code, and open it in the Claude app or at claude.ai/code. Anthropic’s Remote Control docs say you can “attach a photo or file in the Claude app or at claude.ai/code, with or without a caption. Claude sees attached photos directly as part of your message.” It’s how you get a screenshot from your phone, or a photo of a whiteboard sketch, to the session on your desk.

Price: “available on Pro, Max, Ultra, Team, and Enterprise plans. API keys are not supported.” On Team and Enterprise, “an Owner must first enable the Remote Control toggle.”

Watch for: “your computer has to stay on and the claude process has to keep running.” If a file you attach can’t reach your computer, Claude gets a note such as “[1 of 3 attachments did not arrive]” in its place. For Codex, OpenAI’s mobile page covers starting, approving and reviewing tasks from your phone, but doesn’t mention attaching images.

Enjoy

Best for: giving every lab’s agent screenshots the same way, from your computer, a browser or your iPhone.

Enjoy runs Claude Code, Codex, Grok Build, Antigravity, Muse Code, and local models through OpenCode, on Mac, Windows and Linux, on the plans you already have, and its message box takes images the same way for each of them. Attach files and images says to “Drag files from your computer onto the conversation, or paste an image you copied, such as a screenshot,” or use the + button and Add images. “Each attachment appears above the message box with a number, and a matching marker such as [1] goes into your message right where your cursor was, so you can refer to it in your sentence.” It’s the system’s own paste, Command-V on a Mac and Ctrl+V on Windows and Linux, so there’s no key to learn. OpenCode needs a version that supports image attachments.

From the iPhone app, select +, then Take photo or Photo library, and the picture goes to the agent working on your computer. “You can add up to 8 photos, and they’re resized to fit the 2 MB limit.”

Price: Enjoy is one monthly subscription, and the first seven days are free, with a card to start. It includes your phone, any browser and one computer; each other computer costs extra each month, and each teammate you invite needs a subscription of their own. See pricing.

Watch for: “Images can be PNG, JPEG, WebP, or GIF files, up to 20 MB each. A message can include up to 8 images.” In the web app, “each image can be up to 2 MB.” From your phone or a browser, your computer has to be “On and awake,” “Online,” and “Running Enjoy, signed in to your account.” It attaches images only inside Enjoy, not in your terminal or editor. Enjoy isn’t open source, and it’s a monthly subscription on top of your AI plans once the free week ends.

When to stay with what’s built in

If you use one agent in a terminal, you already have what you need: Ctrl+V in Claude Code, or Alt+V on Windows, and a paste or codex -i in Codex. When a paste key does nothing, don’t fight the clipboard: save the screenshot, then drag the file in or type its path. If you’d rather click than remember keys, Claude’s desktop app and the ChatGPT app are included with the plans you have.

If the screenshot is of your own app in a browser, you may not need to take it at all: Claude’s desktop app and the ChatGPT app can open your app and check it themselves, and our browser guide covers the ways to let an agent do that.

Enjoy fits when you run more than one lab’s agent and want one way to hand each of them a picture, or want to send one from your iPhone to any of them.

How we compared

We read Anthropic’s Claude Code docs on common workflows, interactive mode, the desktop app, Linux, VS Code, Remote Control and setup, Claude’s pricing and its API docs on images, OpenAI’s Codex docs on image inputs, the CLI, its command line options, the IDE extension, mobile, the desktop app and pricing, Codex’s source code on GitHub, Apple’s and Microsoft’s guides to screenshots, Terminal and WSL, and Enjoy’s own help on November 18, 2026, and every one is linked below. We quote each company’s own words for its limits and leave out what it hasn’t documented. We didn’t test each key in every terminal, so where a company’s docs are silent, we say so. These apps change weekly, so we recheck this guide against its sources at least once a month, and the date at the foot of the page says when we last did.

13.Questions.

How do I paste an image into Claude Code?

Copy the image, then press Ctrl+V in Claude Code’s prompt. That’s Ctrl+V on a Mac too, since Cmd+V works only in iTerm2, and Alt+V on Windows and WSL. Claude Code puts an [Image

Why won’t Ctrl+V paste an image into Claude Code on Windows?

On Windows, Claude Code’s image paste is Alt+V. In WSL both keys are bound, and Anthropic says to “use Alt+V if your terminal intercepts Ctrl+V.” If neither does, drag the file in, type its path, or use the + button in Claude’s desktop app.

How do I give Codex a screenshot?

In the Codex CLI, paste it into the composer, or start with codex -i screenshot.png and your prompt; for several, separate the paths with commas or repeat --image. In the ChatGPT desktop app and the IDE extension, hold Shift while you drag the image into the composer.

Can I send a screenshot from my phone to Claude Code or Codex?

For Claude Code, yes, with Remote Control, on a paid Claude plan. Attach a photo in the Claude app, and the session on your computer gets it. OpenAI’s page on Codex on mobile doesn’t mention attaching images. Enjoy’s iPhone app sends photos to every lab’s agent on your computer.

Do screenshots use up my Claude Code or Codex limits?

Yes, an image is part of what you send. Anthropic’s API docs say an image “costs ⌈width / 28⌉ × ⌈height / 28⌉ visual tokens”; on Claude 4.7 and later models, a 1920×1080 image is 2,691 tokens, and larger ones are scaled down to at most 4,784. OpenAI doesn’t say how much an image counts in Codex. Either way, crop to the part that matters.

Which image formats work?

Claude takes JPEG, PNG, GIF and WebP, and uses only the first frame of an animation. OpenAI says Codex “accepts common image formats, including PNG and JPEG.” Enjoy attaches PNG, JPEG, WebP and GIF files as images, and other formats, such as SVG, as files.

Enjoy Agentic Industries · Checked against its sources on