Will it just fix my photo for me?

This page doesn't edit anything itself, but it hands your sentence to something that can. Say "edit this PSD for me" and it takes you to the Online PSD Editor page; your exact words are kept, and once you go from there into the editor they're the assistant's first suggestion, one click to start. When it can't match a tool locally, it hands the sentence straight to the AI assistant, which starts right away — it can do far more than these tool pages.

Does it spend my allowance?

Local matching doesn't. The first step is local keyword matching, running in your own browser with zero delay and no network; phrasings like "compress", "remove background" or "add watermark" all stop there. When it can't tell locally and hands the sentence to the AI assistant, that one counts against the AI assistant's own quota — still with no sign-in.

Is what I type sent anywhere?

The ones matched locally don't — the sentence never leaves your browser. When it can't tell locally, your sentence goes to the AI assistant (you'll see it in the editor's chat) — no image goes with it. The ones to tell apart are the 10 in the AI panel (old photo restore, colorize, replace background, ID photo, upscale, beautify, decorate, de-age, split into editable layers, change text in the image): they send the image of your selected layer, and carry a cloud icon next to the title.

Does it know every feature on the site?

What it recognises locally is a table of common phrasings (compress, convert, remove the background, brighten…). Anything not in that table isn't dropped — the whole sentence goes to the AI assistant in the editor, which knows every editor feature, so new features work without waiting for the table to catch up. The editor also holds several hundred menu commands (layer styles, individual filter parameters…); to find one yourself, press Ctrl+K inside the editor.

AI command: say what you want, it finds the tool

You don't have to remember what the tool is called. Say what you want done and let it match.

or start a blank document

Try: make a poster · a logo · remove the background · squeeze to 200KB

Once there are enough tools, the time goes not into the work but into remembering what the feature is called and which menu it hides in. This page turns the entry around: say «make this smaller», «lose the background», «it looks blurry», and it translates that into one of the tools here and sends you over — no need to learn our naming first.

It doesn't touch pixels itself, but your sentence isn't lost. For phrasings it recognises right away (compress, convert, remove the background…), it takes you to that tool's page and keeps your sentence for you — once you're in the editor, the assistant's first suggestion is exactly what you said, one click and it acts. For anything it doesn't recognise, it opens the editor with your sentence as-is and the AI assistant starts on it right away. The one doing the work is the tool or the assistant, not this page.

Key features

Two steps: match locally first, otherwise hand it to the assistant

① Local keywords: phrasings like "compress" or "remove background" match straight away — zero delay, no quota, and almost all real input stops here. ② Not recognized locally: your sentence goes straight into the editor to the AI assistant, which starts right away; every step shows in the chat.

English and Chinese both work

Typing «compress» and typing its Chinese equivalent land on the same tool. The keyword table carries both, rather than expecting you to remember which language the site names things in.

No image — just the sentence

It doesn't carry the image: on a match it only takes you to that tool, and you add the image once you're there. What it carries is your sentence — in the editor it's still one click to run.

It won't pretend to have matched

When it can't tell locally, it won't just pick the closest-looking tool and send you there — the wrong place wastes more time than saying "I don't know". It hands your sentence straight to the AI assistant: into the editor, and the assistant starts right away — it can do far more than these tool pages.

How to use

1
No need to upload first

The command bar doesn't carry an image; it only takes you to the tool, and you add the image once you're there. The drop area above is a separate route: drag an image into it and it goes straight into the editor with it.
2
Say it in your own words

«too big», «remove the background», «the photo looks yellow», «I need a 3×3 grid» — never mind what we called the tool.
3
See what it picked

If it recognizes the sentence it shows the tool's name and takes you there; if not, it takes you into the editor and the AI assistant carries on. Local matching costs no quota.
4
Pick it up on the other side

Just do it right there on the tool page; to hand it to the assistant, go from that page into the editor: on most tool pages that produce a result right there, the toolbar shows “Pro editor” once you've added an image (on narrow screens some pages show just a diagonal double-arrow icon); the collage, long-image, GIF, video-frame and three batch pages don't have that button, so click “or start a blank document” under the drop zone to get into the editor; the other tool pages take you straight into the editor when you drop an image in. Your original sentence is the assistant's first suggestion — one click and it runs.

FAQ

Will it just fix my photo for me?: This page doesn't edit anything itself, but it hands your sentence to something that can. Say "edit this PSD for me" and it takes you to the Online PSD Editor page; your exact words are kept, and once you go from there into the editor they're the assistant's first suggestion, one click to start. When it can't match a tool locally, it hands the sentence straight to the AI assistant, which starts right away — it can do far more than these tool pages.
Does it spend my allowance?: Local matching doesn't. The first step is local keyword matching, running in your own browser with zero delay and no network; phrasings like "compress", "remove background" or "add watermark" all stop there. When it can't tell locally and hands the sentence to the AI assistant, that one counts against the AI assistant's own quota — still with no sign-in.
Is what I type sent anywhere?: The ones matched locally don't — the sentence never leaves your browser. When it can't tell locally, your sentence goes to the AI assistant (you'll see it in the editor's chat) — no image goes with it. The ones to tell apart are the 10 in the AI panel (old photo restore, colorize, replace background, ID photo, upscale, beautify, decorate, de-age, split into editable layers, change text in the image): they send the image of your selected layer, and carry a cloud icon next to the title.
Does it know every feature on the site?: What it recognises locally is a table of common phrasings (compress, convert, remove the background, brighten…). Anything not in that table isn't dropped — the whole sentence goes to the AI assistant in the editor, which knows every editor feature, so new features work without waiting for the table to catch up. The editor also holds several hundred menu commands (layer styles, individual filter parameters…); to find one yourself, press Ctrl+K inside the editor.

Try AI command: say what you want, it finds the tool now

No install · no sign-up · local stays local · free AI credits every day