Bundle your errors and 100+ project files into one paste-ready block for Claude, ChatGPT, or any AI — then paste the reply back and apply the edits to your files. Full round trip, inside your editor.
Open VS Code, press one shortcut, paste into any AI chat. That's the entire workflow.
Open any project in VS Code. SendToAI lives in the sidebar and auto-scans your file tree — ignoring node_modules, lock files, binaries, and anything in .gitignore.
Press the keyboard shortcut to bundle. Select exactly which files to include with the visual file picker, choose your format, and add a prompt template.
Ctrl+Alt+A ⌘+Alt+AYour entire project is on the clipboard. Paste into Claude, ChatGPT, or any AI chat — then copy the reply and hit Apply AI Reply to write the edits back to your files. Reviewable, undoable with Ctrl+Z.
One click bundles every file VS Code is flagging plus the exact error messages, with a ready-to-paste "please fix these" prompt. Web bundlers can't see your Problems panel — this can.
Copy the AI's answer, run one command, and pick which edits to write back to your files. Reviewable in git diff, undoable with a single Ctrl+Z. No more hand-copying snippets.
When a bundle must shrink to fit, files are ranked by relevance to your prompt — and files they import survive the cut with them. Connected context, not orphaned snippets.
Checkbox tree with tri-state directory selection. See token estimates per file and folder before you bundle. Select exactly what the AI needs.
Save your file selections as named presets. One click to reload "backend only", "frontend only", or any custom context you use repeatedly.
Bundle only modified and staged files from your current git diff. Perfect for asking AI to review or explain just what changed.
Select a target AI (ChatGPT Free 8K, ChatGPT Plus 32K, Claude 200K, Gemini 1M) and SendToAI keeps what actually matters for your task — ranking every file by your prompt and following the import graph so its dependencies survive the cut too. Dumb bundlers drop the biggest files; this drops the least relevant. Never crash your AI with an oversized paste again.
After every bundle, see exactly how full you are against ChatGPT Free, ChatGPT Plus, Claude, and Gemini — color coded green/yellow/red so you know exactly what will work.
Standard plain text, Claude XML (structured tags), Compact (strips comments, ~20% fewer tokens), or Minimal (max token savings). Pick what works for your AI.
Persistent context memory per project. Save architecture notes, tech stack info, or key constraints and include them in every bundle automatically.
Bundle only the files currently open in your editor. Great for focused sessions where you're only working in a few files.
Your source code never leaves your machine. All bundling is local. We collect zero telemetry, zero analytics, zero file data. Ever.
The free tier is genuinely useful. Pro unlocks the power features for heavy users.
| Feature | Free | Pro |
|---|---|---|
| Files per bundle | Up to 50 | Unlimited |
| Bundle Errors (from the Problems panel) | ✓ | ✓ |
| Apply AI Reply (write edits back, undoable) | ✓ | ✓ |
| Output formats (Standard, XML, Compact, Minimal) | ✓ | ✓ |
| Open Tabs mode | ✓ | ✓ |
| Visual file picker | ✓ | ✓ |
| Token + cost estimator | ✓ | ✓ |
| 4 context window bars | ✓ | ✓ |
| Project notes / context memory | ✓ | ✓ |
| Named presets | — | ✓ |
| Git diff mode | — | ✓ |
| Fit to AI Window (smart relevance trim) | — | ✓ |
| Pro license machines | — | 3 |
| Price | Free | $9 one-time |
Install free in 10 seconds. Works with every AI. Your code stays on your machine.