Cleanup.pictures

What is Cleanup.pictures?

Cleanup.pictures is a browser-based AI inpainting tool for removing unwanted things from photos. You paint over an object, a person, a line of text or a blemish, and the model fills the masked area with a reconstruction of what plausibly sat behind it. The product does one retouching job, and its pricing is built around how large the exported result may be.

The service belongs to the Clipdrop family of image tools. Init ML's general terms apply to the Clipdrop and Cleanup.pictures websites, so the same contract covers both sites. The company behind the service is INIT ML, a French simplified joint-stock company.

Ownership has changed since launch. Jasper acquired Clipdrop from Stability AI for an undisclosed price, less than a year after Stability AI had bought Clipdrop through its parent brand, Init ML. The current Cleanup API documentation reflects that change: Clipdrop is now part of Jasper, and developers are directed to Jasper's team for newer image models and credits.

At launch in 2021, a design-media report said the app was built with the help of LaMa, an open-source model created by Samsung. The current homepage and API documentation do not name the model in use today, so that launch-era detail describes the original build rather than a confirmed current engine.

Core features

• Object and person removal: Cleanup.pictures removes unwanted objects, people or defects from photos, which makes it a direct way to remove objects from photos without manual cloning.
• One-stroke reconstruction: the algorithm reconstructs what was behind an object in one click, as long as the unwanted element is fully covered; people and text are removed the same way.
• Text, logo and watermark removal: it can also remove unwanted text, logos, date stamps or watermarks, subject to the licensing warning described under Limitations.
• Blemish and wrinkle retouching: the CleanUp brush can remove blemishes or wrinkles from a profile picture; you overflow the brush slightly past the spot and download the result.
• Quality modes: on the API side, the default fast mode matches the default mode on the Clipdrop website, while the slower quality mode matches the HD mode.

The vendor's main technical claim is about how gaps get filled. It contrasts its AI with clone-stamp tools that need a background reference, saying its model can guess what was behind unwanted text, people or objects in a few clicks. In practice, the workflow is to mark what should disappear rather than to pick a source area to copy from.

The feature set is narrow. Removing an entire background is outside the editor's job: for that task, the FAQ sends users to ClipDrop rather than to Cleanup.pictures itself. If you need to remove people from photos, erase a timestamp or patch a crack, Cleanup.pictures covers it; if you need cut-outs, new backdrops or generated scenery, the sibling Clipdrop tools listed under Cleanup.pictures alternatives are the intended route.

Guide

Removing an object, person or watermark

1. Open the Cleanup.pictures homepage and drag and drop an image onto the page; the editor opens without payment.
2. Pick a brush size. The FAQ recommends choosing a bigger brush and covering more than the area you want to retouch, especially to include shadows.
3. Draw over the object, person, text or watermark. After a few seconds, the marked element should be gone.
4. Download the result. On the free tier the export is resolution-capped, as explained under Pricing and Limitations.

Getting cleaner results

• Overpaint rather than trace: the same overflow advice appears for objects, text and blemishes; the FAQ says covering slightly more than the target, shadows included, helps the algorithm produce the best result.
• Automatic masks in the API: for automatically generated masks, the developer documentation recommends expanding the detected object mask by about 15% to get the best results.

Use cases and examples

The vendor's own examples cluster around small, well-defined removals in otherwise usable photos:

• Photographers: photographers use it to remove time stamps or tourists from holiday pictures before printing them for their customers.
• Old or damaged prints: Cleanup.pictures is also positioned for removing cracks on photographs.
• Real estate listings: for property photos, it is positioned to depersonalize and clean pictures of any room, flat, house or apartment.
• E-commerce and social posts: online sellers are pitched on creating the ideal product shot and updating social media, including visuals for Instagram stories.

All of these examples come from the vendor's use-case copy rather than independent case studies. The free export cap and the compatibility caveat in the terms apply to each of them, which matters most for print work and listing photos that need full resolution.

Who is it for

• Photographers and photo retouchers who need a quick fix for distractions, time stamps or cracks without opening a full desktop editor.
• Creative teams: creative users are pitched on remixing existing photos by replacing parts with their own material.
• Individual subscribers: each Cleanup subscription is individual and limited to one user, which matters for studios that split retouching across several people.
• Developers: teams that need inpainting inside their own product are the API audience; the Cleanup API is described as usable from environments such as Node.js, SwiftUI and Kotlin.

It is a weaker fit for anyone who mainly needs background removal or replacement, which the FAQ hands off to ClipDrop. It is also a poor match for confidential material, given the retention terms described under Privacy and data handling.

Platforms

• Web editor: Cleanup.pictures is used through its website, and a subscription works on both mobile and desktop.
• Device limit: the terms limit use to up to 2 devices per account, specifically 1 mobile and 1 desktop.
• API requests: Cleanup API calls are HTTP POST requests with a multipart/form-data body carrying two image files (the original and a mask) plus an optional mode field.
• API input limits: the original image must be a JPG or PNG of up to 16 megapixels and 30 Mb. The mask must be a PNG at the same resolution, up to 30 Mb, in black and white with no grey pixels.
• API output: a successful call returns a PNG image with the same dimensions as the original image.
• Rate limit: each API key is limited by default to 60 requests per minute on the cleanup API, and users can ask the company for higher values.

The web editor and the API are separate entry points. A Pro subscription covers the browser editor, while API usage runs on its own credits and keys, described under Pricing.

Pricing

Cleanup.pictures is free unless you need better quality and high-resolution processing, which is exactly what the paid tier sells. Pro is listed at $5 per month or $36 per year ($3 per month) for processing images of any size, and the trial lets you test HD quality for free. The trial length is not stated on the homepage. The Pro column also lists unlimited resolution and a high quality refiner.

API access is priced separately, and the homepage notes that special pricing applies to the API. Logged-in developers can claim 100 free Clipdrop API credits for development and debugging, and further calls are rejected once those 100 images are used.

Renewal, cancellation and refunds

• Automatic renewal: subscriptions renew tacitly for periods of the same length as the initial monthly or annual period.
• Notice period: the terms require cancelling at least 7 days before the end of a monthly period and 1 month before the end of an annual period.
• Managing the plan: the FAQ says a subscription can be edited, paused or cancelled from a "manage subscription" section.
• Started periods: once the services end, account access ends too, and any period that has already started is due in its entirety.
• Refund conditions: after the license is activated, refunds are given only in the rarest cases, such as technical difficulties, platform incompatibilities or other unforeseen circumstances.
• Refund amount: when a refund does apply, it is total if the subscription is less than 14 days old and partial, prorated by days, if it is older.

Taken together, the trial is the intended evaluation window. An annual plan needs a month's notice before renewal, and refunds after activation are described as exceptional.

Cleanup.pictures alternatives

• ClipDrop Pro (same company): the Cleanup.pictures pricing table itself includes a ClipDrop Pro column that bundles Cleanup with background removal, background replacement, uncrop and generative fill.
• Clipdrop's own Cleanup tool: on Clipdrop's pricing page, the Free tier lists Cleanup at 20 uses per 24 hours, whereas the Cleanup.pictures free column lists unlimited images at a capped resolution.
• Clipdrop Remove Background: for background work, the homepage promotes a separate Remove Background tool that downloads images with a transparent or white background.
• Adobe Photoshop: Cleanup.pictures positions itself against complex software such as Adobe Photoshop, claiming professional results in a few clicks.
• General online editors: a third-party AI tool directory lists Canva, Fotor and PhotoRoom among the top Cleanup.pictures alternatives, each described there as a broader product: a template-based design tool, an all-in-one photo editor and generator, and an AI editor for product images.

The choice mostly comes down to scope. Cleanup.pictures is the single-purpose removal brush with a cheap Pro tier; the Clipdrop bundle adds background and generative tools from the same company; Photoshop is the complex software the vendor positions itself against; and the general editors fold object removal into a wider design or product-photo workflow.

Limitations

Output and quality limits

• Free export cap, stated two ways: the pricing table describes the free tier as unlimited images with resolution limited to 720p. The FAQ instead says images of any size can be imported and edited, but free exports are limited to 720px, while Pro has no size limit. The two statements on the homepage use different units, so the exact free cap depends on which one applies.
• Compatibility: the terms state that certain photographs or images of objects may not be compatible with the service, and the company does not accept responsibility for defects in the quality of the montage or photos.
• Speed versus quality: in the API, the default fast mode may produce artifacts in the result image, while quality mode is slower but produces better results.

Rights and acceptable use

• Watermarks: the vendor warns that watermarks usually indicate restrictive copyrights and says to remove watermarks only on images for which you hold an explicit license.
• No dataset building: the terms prohibit using the services for research or development of artificial intelligence, in particular the generation of training data.

Support and refund contacts

• Refund address: the refund contact address printed in the terms misspells the Clipdrop domain with a double l.
• Technical support: technical assistance is offered by email at a separate address on the correctly spelled Clipdrop domain, so the two contacts in the same document do not match.

Privacy and data handling

• Policy scope: the privacy policy reached from the Cleanup.pictures footer is written for the ClipDrop application. It shows an entry-into-force date of June 21 2022.
• Processing purposes: to provide its services, the policy lists analysing and processing images, transferring them between devices, and storing users' photos on its cloud for service improvement purposes. The policy text does not mention model training either way.
• Retention: without an account, images, photographs and texts are retained for two years.
• Hosting: personal data are hosted on Google Cloud Platform servers located in the United States.
• Responsibility for sensitive images: users accept that personal or confidential content is "downloaded, stored, and carried out" under their sole responsibility.
• Deletion rights: under the GDPR, users have the right to request that their personal data be deleted and to prohibit any future collection.

FAQ

Q1. Is Cleanup.pictures free?

Yes, for basic use. The free tier covers unlimited images, but exports are capped: the pricing table says 720p and the FAQ says 720px. Full-resolution export and the high quality refiner require Pro.

Q2. How much does Cleanup.pictures Pro cost?

Pro is listed at $5 per month or $36 per year, which works out to $3 per month on the annual plan. Subscriptions renew automatically, and cancelling an annual plan requires one month's notice before the period ends.

Q3. Can Cleanup.pictures remove a photo's background?

Not as its main job. The FAQ points background removal to ClipDrop, and the ClipDrop Pro bundle shown on the Cleanup.pictures pricing table adds background removal, background replacement, uncrop and generative fill.

Q4. Is there a Cleanup.pictures API?

Yes. The Cleanup API takes an original image and a black-and-white mask in one POST request, returns a PNG at the original dimensions, and allows 60 requests per minute per key by default. Developers can claim 100 free credits for testing, and further credits go through Jasper's team.

Q5. Can I use Cleanup.pictures to remove watermarks?

Technically yes, since the brush removes text, logos and watermarks. The vendor warns that watermarks usually signal restrictive copyrights and should only be removed from images you have an explicit license for.

Q6. Who runs Cleanup.pictures?

INIT ML, a French company whose general terms cover both the Clipdrop and Cleanup.pictures websites. Clipdrop was bought by Stability AI and later acquired by Jasper, and the API documentation now describes Clipdrop as part of Jasper.

Q7. What happens to photos uploaded to Cleanup.pictures?

The privacy policy says images, photographs and texts from users without an account are kept for two years, data are hosted on Google Cloud Platform in the United States, and photos may be stored for service improvement purposes.