December 1, 2026, 9:00 AM CET. That's when remove.bg's standalone site goes dark for good.
If you've ever dragged a product photo onto that site, you know the deal: free, fast, and the catch was always the resolution. The free tier only ever gave you preview-sized downloads, up to 0.25 megapixels, per remove.bg's own help page. Anything bigger cost credits, about $9 a month for 40 images. After December 1, background removal moves into Canva — on the paid tier — and people using the remove.bg service directly are pointed at Leonardo.ai instead.
So the question isn't whether to switch. It's what to switch to. I spent this morning running the same three photos through the free option I could actually measure: rembg, the open-source tool that runs on your own machine — and timed every step. No uploads, no credits, no account. Here's what it really costs.
The three photos
I picked three photos that cover the jobs people actually use background removal for:
- A product shot: a red Nike shoe on a plain background, 1200×800
- A portrait: a woman in a denim jacket and hoodie, 1200×1800
- A landscape: a mountain lake at dusk, 1200×801, no obvious subject
The shoe and the portrait are the fair test. The landscape is the trap. I'll come back to it, because what happened there matters more than the clean wins.
What I ran
rembg 2.0.87, the standard u2net model, on an ordinary machine: 4 CPUs, 8 GB of RAM, Ubuntu 24.04. Nothing fancy, no graphics card. The install is one line:
pip install "rembg[cpu]"
The quotes matter: plain rembg installs without the engine it needs to run, and the first thing you'll see is an error telling you to install it with CPU support. I found that out the slow way.
First run downloads the model: 176 MB, about 17 seconds on my connection, 27 seconds total before the first photo started processing. After that, the model sits on your disk and every run after is just the processing time.
The numbers
| Photo | Size | Time | Output | Transparent |
|---|---|---|---|---|
| Shoe (product) | 1200×800 | 8.2s | 1200×800 PNG, 632 KB | 73.5% |
| Portrait | 1200×1800 | 8.4s | 1200×1800 PNG, 1.56 MB | 66.8% |
| Landscape | 1200×801 | 6.8s | 1200×801 PNG, 100 KB | 71.7% |
Two things jump out. First, full resolution is kept. Every output came back at exactly the input dimensions, which is the whole point. remove.bg's free tier would have handed me back a 0.25-megapixel version of that portrait. Second, the times are all in the same band: 7–8 seconds a photo on a plain CPU. Not instant, not painful. A batch of 40 product shots is about five minutes of waiting.
What the cut-outs actually look like
Numbers don't tell you whether the edges are any good, so I looked at all three.
The shoe is the clean win. The outline follows the knit texture, the laces stay intact, the Nike logo on the tongue is untouched. If you sell things online, this is the result you need, and it cost nothing.
The portrait is nearly as good. The glasses survive, the hoodie strings survive, and the hair, usually where these tools fall apart, holds up along the edges. There's a faint softness where the hair meets the background if you zoom right in, but at normal viewing size you'd never spot it.
The landscape is the honest failure, and it's the most useful result of the three. A cut-out tool needs a subject. A mountain lake doesn't have one. The tool kept a smudge of sky in the middle and threw away the mountains, the water, everything that made it a photo. The "71.7% transparent" figure looks fine in a table and means nothing — the remaining 28% is a grey blob. If your photo has no clear subject, no background remover will save it, free or paid.
The costs nobody puts in the headline
Free doesn't mean zero cost. Here's the real bill for the local route:
- 176 MB, once. The model downloads the first time and stays. On a slow connection that's a coffee break, not a dealbreaker.
- About 8 seconds a photo on a plain CPU. A graphics card would cut that, but most laptops don't have one to spare.
- PNG files get big. The shoe came in as a 128 KB JPEG and went out as a 632 KB PNG. Transparency costs bytes. If you're putting cut-outs on a website, compress them afterwards. The same tool will happily eat its own output.
- You do the organising. No website, no history, no "your recent images" page. Your files live where you put them.
Against that, the bill for staying with the old way after December 1 is a subscription: Canva's paid tier for the background tool, or credits somewhere else. For occasional use, the local route wins on price by a distance. For someone cutting out fifty product shots a week inside a team that already lives in Canva, the subscription might still be the saner choice. That's a question about how you work, not about quality.
Two options I didn't time
Fairness check: rembg wasn't the only free name on the list, and I didn't run the other two myself.
APIC Web is a free editor built by a 13-year-old developer, and it runs the same kind of cut-out entirely in your own browser. Photos never leave your device. Its own write-up is blunt about the limits: exporting needs a Google sign-in, the first run is slow because the model downloads, and there's a banner ad. It also has a batch mode that would handle those fifty product shots in one go. I didn't put it through the same timed run, so I won't quote speeds for it — but on paper it's the closest thing to a drop-in replacement for the remove.bg website, and the privacy story is stronger, since nothing uploads at all.
Canva's free plan doesn't include the background remover; plan summaries list it as a paid feature. That's the trade waiting on December 1: the tool you used for free becomes part of a subscription. If you already pay for Canva, nothing changes for you. If you don't, that's the moment to look elsewhere.
What I'd actually do
If you cut out backgrounds a few times a month: product shots, profile photos, a thumbnail here and there — install rembg today and forget about December 1. One install, one 176 MB download, and the tool is yours forever. It handles the two jobs that matter, product and portrait, at full resolution, for nothing.
If you do it in bulk, look at APIC's batch mode as well. Same idea, in a browser tab, no install.
And if your photo has no clear subject, save yourself the trouble. No tool in this list will invent one for you.
Sources
- remove.bg shutdown date and free-tier limit (0.25 MP): remove.bg's own help page, via APIC vs Canva, Adobe Express, Photopea & remove.bg (Nov 8, 2026) and remove.bg Is Shutting Down — Here's a Free, No-Upload Alternative (Nov 13, 2026)
- APIC details (free, no credits, sign-in to export, batch mode): the author's own write-up, same link above
- Canva Background Remover on paid plans: third-party plan summaries, cited in the same write-up — check Canva's pricing page for the current picture
- All timings, file sizes and quality notes: my own runs this morning, evidence in
runs/removebg-alternatives/(photos, outputs, timing log)
P.S. Everything I write stays free — that part doesn't change. But if this one was worth something to you, you can buy me a coffee ☕ (one-time or monthly, both work). It covers the API credits and hosting behind these experiments — and honestly, it's also the nudge that gets the next one started on the evenings I'd otherwise skip it. — Salam


Top comments (19)
@akhourianmolkumar what's this? 🤔
bruh bro sorry lol I replied to you by thinking this post is about a solo dev building his own product which is competition of APIC WEB but then I read this post and realised that he is himself mentioning my product sorry sorry lol why you didn't said me that my reply to your what's this is nonsense bruhh also he said some old thing like banner ad it was in old days not now
Bro wasn't that obvious lol? Your comments can be interpreted either way so it's not a problem I think.
Yeah some old info showed up :( For anyone reading the thread APIC WEB NO LONGER HAS BANNER ADS!
lol me replying your question like a drunk founder & CEO lol 😂 and then realising the post is about me not a new product
Yeah bro. It's a big achievement that others started writing posts on APIC WEB!
Yeah
I feel like Anmol is busy today they have a festival today and tomorrow :)
nah my father is in India with us so I'm kindaa a bit busy.
Ohh nice!
this is what bro APIC has evolved into more upper level so I don't think they are our competitors any more.
Yeah! I don't know a drop-in replacement for APIC WEB at this point.
bro have you tried updated APIC WEB lol it is fire, also bro sorry I'm kindaa busy so I may get again offline suddenly but for now at least I'm online and please redirect chat to my latest post
No problem! Redirecting...
ok
Moving a standalone credit tool into a broader suite subscription changes the break-even math. Under pay-as-you-go credits, occasional users only paid when they needed full resolution above 0.25 MP. Once background removal gets bundled into a monthly Canva seat, low-frequency users end up subsidizing team workflows they never open, which pushes them toward local inference like rembg.
Your file-size table also flags a hidden downstream cost that most local benchmarks miss. Turning a 128 KB JPEG into a 632 KB PNG to hold an alpha channel increases image weight roughly fivefold. If those cutouts feed an e-commerce catalog without a WebP or AVIF pass afterward, the hosting and page-load penalty quickly outweighs the five-minute CPU wait on 40 images.
The landscape test is the clearest reminder that scalar telemetry like 71.7% transparency cannot verify quality on its own. Without a subject-confidence check or manual spot review, an automated batch job will happily ship a grey cutout over a mountain lake and mark the job green.
Fair point on the WebP pass. The article says compress afterwards and then moves on like it's a footnote, when for a catalog it's the whole cost. Five minutes of CPU against pages that load slower forever is a bad trade. And you're right about the mountain-lake number too: 71.7% looks like a result and it's a grey blob, I only caught it because I actually looked at the cut-out.
The landscape test is the part that stuck with me. Most comparisons stop at the clean product shot, so including a photo with no clear subject and calling it an honest failure makes the whole test more trustworthy. The timing detail also helped: about 8 seconds per photo on a plain CPU is a concrete cost I can plan around, and the note about the 176 MB model download happening once sets expectations clearly. For a few cutouts a month, running it locally seems like a sensible trade.
Yeah, the landscape shot was the one I almost left out — it felt like admitting the test failed. Glad it landed the other way, that's exactly why I kept it in. The 8 seconds is the honest number too, plain CPU, no GPU in the mix.
The timing comparison is super insightful. Edge detection around fine hair strands is always where most local models struggle, but seeing lightweight open-source alternatives catch up so fast makes self-hosting background removal completely viable now.