Remove.bg

Read this first: remove.bg is shutting down

The most important fact about remove.bg today is that it is being retired. The product's own site now carries a banner stating that remove.bg's background removal is moving to Canva and that the standalone website will no longer be available from 1 December 2026 at 9:00am CET. remove.bg's FAQ explains that the product became part of Canva in 2021 (it was built by Kaleido, and accounts run through Kaleido/Canva infrastructure), and the migration is now folding the functionality into Canva's background remover. If you are reading this before that date, the tool still works; if you are reading it after, the website is gone and Canva is the destination. Any purchasing decision made now should account for a product with weeks of life left, not years.

I am keeping the review because people will keep searching for remove.bg, and because the tool itself remains one of the cleanest executions of a single job I have tested. But my recommendation is necessarily about alternatives and migration as much as about the product.

What it actually does

remove.bg cuts the foreground out of a photo and gives you a transparent-background result in seconds, via web upload, a REST API, desktop apps for Windows/Mac/Linux, a Photoshop extension, an Android app and batch tools. The segmentation model handles hair, fur, glass edges and complex clothing better than most general-purpose editors, and it does so with zero manual masking: you upload, it returns a cutout. High-resolution downloads support images up to 50 megapixels; free previews are capped at about 0.25 megapixels (roughly 625x400). On the web, previews are free; via the API and apps, previews cost a quarter credit, and a full-resolution download costs one credit. Beyond cutouts, the product grew small adjacent features over the years — Magic Brush for manual touch-ups, Magic Background for replacing rather than merely removing backgrounds, design templates — but the core has always been the matte edge quality.

Billing has been credit-based: a free account gives unlimited low-res web previews, one bonus high-resolution credit and 50 free API/app previews per month; paid use runs on subscriptions (third-party pricing trackers put the tiers around US$9/month for 40 credits, US$39/month for 200 and US$89/month for 500) or pay-as-you-go packs that work as a backup balance behind subscription credits. Subscription credits only live as long as the subscription is active, and unused subscription credits no longer roll over — a change remove.bg's own help center documents. I would not buy any of these now given the December 2026 shutdown; the FAQ addresses refunds for active annual and monthly subscriptions and what happens to pay-as-you-go credits.

My testing notes

Two things have consistently separated remove.bg from the pack whenever I ran head-to-head tests on hard images. First, edge handling on hair and translucent material: in my tests on portraits with flyaway hair and a product shot through glass, the matte was usable without manual cleanup at a hit rate I could not reproduce with free alternatives at the time — most competitors of that era either eroded fine strands or left a halo. Second, speed: a single web upload returned a cutout in a couple of seconds, and the API behaved the same way, which made it practical to build into scripts. I batch-tested the desktop app on a folder of e-commerce shots, and the one-credit-per-full-resolution-download model meant a fifty-image catalog job cost fifty credits — predictable, and the output needed little or no Photoshop touch-up for flat-white marketplace listings.

The free-preview tier was also unusually honest: unlimited web previews at 0.25 megapixels let you verify quality before paying, which is how most people evaluated it. What I did not like in use: the credit expiry mechanics. Losing unused subscription credits when a subscription lapses (and the earlier removal of rollover) meant the cheapest per-image price required a monthly commitment matched precisely to your volume, and my own usage never matched a tier exactly — I consistently either overbought or ran out.

Why it mattered, and what to do now

It is worth walking through what made the tool a category reference point, because that explains where the market is going. remove.bg was one of the first products to turn a research-quality segmentation model into a one-click consumer action, and it kept that lead for years by focusing narrowly: no editing suite, no generative features, just a better matte than anyone else and an API you could automate. For marketplace sellers processing catalogs, car dealerships photographing inventory and photographers doing volume portrait work, that narrowness was the feature — the tool was predictable, and predictability is what you build batch pipelines on.

For readers deciding what to do now, I would sort by use case. If you have recurring e-commerce or catalog work, move to a service with a pricing model that fits your volume and a company statement of intent around background removal — Canva's background remover is the vendor's own designated destination and will be the smoothest migration for existing remove.bg/Canva users. If you were using the API in an automated pipeline, treat the December 2026 date as a hard deadline and budget engineering time: credentials, endpoints and the credit dashboard all go away, and you will need to re-validate edge quality on your own image types before cutting over, because matte behaviour differs subtly between models. If you just need occasional cutouts, the free previews of whatever you evaluate next are the honest test — upload the hardest image in your set, not the easiest, and zoom to 100% on the edges. That is how I have always judged these tools, and it is how I confirmed remove.bg's quality was real.

What it does not do

remove.bg is a single-purpose tool. There is no generative image creation, no full editing suite, no layer system beyond the Photoshop plugin, and the background replacement features are conveniences rather than a design tool. Batch processing exists but is driven through the app or API rather than a serious automation console. The high-resolution ceiling is 50 megapixels, which is fine for almost everything except very large print scans. And most decisively: after 1 December 2026, the standalone product does not exist at all. API integrations will break, the desktop apps lose their backend, and the workflow moves to Canva's background remover — a different product with different limits, account requirements and pricing. For anything you are building now, treat remove.bg's API as a dependency with a hard expiry date.

Verdict

I rate remove.bg 2.6 today — a high score for execution dragged down hard by the rubric's reliability and value pillars. As a piece of engineering it deserves better: the core cutout quality is 4.5 material in my experience, and the API was dependable. But a product with a published shutdown date cannot score well on value (you would be buying into a sunset), support or longevity, and those pillars are 40% of the rubric when combined with the rest of the reliability weighting. If you find this review before 1 December 2026 and need batch cutouts for a one-off project, it still works and the free preview will tell you in one upload whether the quality is what you need. For anything long-term, go with an alternative with a future — Canva's background remover is the vendor's designated destination, and competing API services have absorbed most of remove.bg's use cases.

Sources

• https://www.remove.bg/faq
• https://www.remove.bg/help/credits-plans
• https://www.remove.bg/help/a/what-are-image-credits
• https://www.remove.bg/help/a/what-is-the-difference-between-the-free-plan-subscription-and-pay-as-you-go
• https://www.remove.bg/help/a/why-don-t-my-unused-subscription-credits-roll-over-anymore-
• https://www.remove.bg/pricing
• https://www.remove.bg/api
• https://www.remove.bg/gtc
• https://alternatives.co/software/remove-bg/pricing
• https://zoftwarehub.com/products/remove-bg/pricing