remove.bg

Give agents one-call background removal through remove.bg. Strip backgrounds from product shots, portraits, and car photos, returning transparent images ready for listings, ads, and design pipelines.

2 actions API key available

remove.bg website ↗ Platform Documentation ↗ Capability data verified 2026-07-27

A folder of new product photos lands in cloud storage from the shoot

Get Account checks the remaining credit balance and the resolution sizes the plan allows against the size of the batch

Remove Background processes a small sample with the product foreground type, cropped to subject, PNG with transparency

The listings owner reviews the sample cut-outs for edge quality, fine detail, and shadow handling before the batch is allowed to spend

Remove Background runs the approved batch with the same settings, saving each cut-out to durable FlowRunner file storage

The agent compares creditsCharged across the run against the pre-run balance

Finished transparent images upload back to the listing pipeline, named to match their source files

The channel receives the batch summary: images processed, credits consumed, balance remaining

What This Integration Enables

Background removal is the least glamorous bottleneck in commerce content. Marketplaces want clean subjects on white, ad sets want transparent cut-outs, design pipelines want alpha masks, and every one of those images used to cross a human desk. remove.bg turned that craft into a single API call, and FlowRunner turns the call into a governed pipeline step: images flow in from storage, storefronts, and shoots, cut-outs flow out to listings and compositing, and the two constraints that actually matter in production, credits and quality, are managed instead of discovered. - Strip backgrounds from product shots, portraits, and vehicle photos with foreground-type detection tuned per subject - Replace backgrounds with a solid color or a replacement image in bulk, for marketplace and campaign consistency - Produce alpha masks for downstream editing and layering in design tools - Control output precisely: format, resolution, region of interest, crop to subject, scale, position, and drop shadow - Check the credit balance and plan limits before large batches, and reconcile creditsCharged after Each successful call consumes credits from a finite balance, which is why batch work here runs behind a [human-in-the-loop](/concepts/human-in-the-loop) gate rather than on a loop.

Without FlowRunner

Cut-outs are a production queue Product photos wait on someone with an editor and a pen tool, one image at a time

Batch spend is discovered afterward A big run burns through credits, and the balance surprise arrives with the invoice

Quality checked at the end Edge problems surface after the whole batch is processed, when the credits are already gone

With FlowRunner

Background removal is a pipeline step New images are cut out automatically on arrival, in the format the listing flow needs

Every batch is priced before it runs Credit balance and plan resolution are checked up front, and the spend is approved, not discovered

Quality gates the batch, not the postmortem A reviewed sample releases the run, so a systematic failure costs a handful of credits

Use Case Scenarios

Product photos become listing assets on arrival

A photographer drops the shoot into [Dropbox](/integrations/dropbox-service). Each new file triggers the flow: the agent gets a temporary link, passes it to Remove Background with the product foreground type and crop-to-subject enabled, and files the transparent PNG back into the listings folder. By the time the merchandiser opens the folder, the raw shots have clean counterparts ready for the storefront, named to match.

Generated subjects, composited campaigns

The creative flow generates a subject with the [AI Image Generator](/integrations/ai-image-generator), then calls Remove Background to produce a transparent cut-out ready for compositing onto campaign backgrounds. Variants for different placements come from the same call: a solid brand-color background for the catalog tile, an alpha mask for the designer working in layers. One subject becomes a placement-ready set without an editor touching it.

The archive cleanup priced before it runs

Three years of product photos in [Google Drive](/integrations/google-drive-api) need consistent white backgrounds for a marketplace migration. The agent inventories the folder, calls Get Account, and reports the arithmetic: this many images, this balance, this plan resolution. The owner approves the spend, a sample of representative images goes through review first, and the batch runs in paced chunks with the balance checked between chunks. The migration finishes with no surprise on the statement.

Human-in-Loop Highlight

The costly operation in this connector is Remove Background at batch scale, because the API charges credits for successful calls, not for good ones. A cut-out that technically succeeded but clipped the product's fine detail, swallowed a strap, or mangled hair still consumed its credit, and a systematic mismatch, the wrong foreground type against a whole catalog, multiplies that waste across every image in the run. So FlowRunner batches spend nothing at scale until a person has seen the evidence. The agent processes a representative sample, presents the cut-outs beside the originals with the credit math from Get Account, and waits. Approval releases the batch with the proven settings; a rejected sample costs a handful of credits instead of the balance. The machine does the cutting; the person who answers for the content budget decides when it runs a thousand times.

Agent processes routinely

Detects exception requiring judgment

Clear match Continues automatically

Ambiguous Routes to human via preferred channel

Human decides

Agent resumes with decision

Agent Capabilities

2 actions

Background Removal

Remove Background Removes an image's background with remove.bg's AI and saves the cut-out to FlowRunner file storage, returning a durable file URL. Accepts exactly one source: a file URL, a public image URL, or base64 image data. Supports foreground type detection for person, product, car, animal, graphic, and transportation subjects, output size and format control (transparent PNG, JPG, auto, or a ZIP of color image plus alpha), alpha-mask-only output, solid color or image replacement backgrounds, drop shadows, semi-transparency handling, region of interest, crop to subject, scaling, and positioning. Each successful call consumes credits, reported back in creditsCharged when the API exposes it.

Account

Get Account Retrieves the account's credit balance and API usage: total available credits with the subscription, pay-as-you-go, and enterprise breakdown, free preview calls remaining this month, and the resolution sizes the plan allows. The check that runs before every batch and the reconciliation source after it.

Frequently Asked Questions

What can FlowRunner do with remove.bg?

FlowRunner agents can run Remove Background and Get Account in remove.bg.

Does connecting remove.bg to FlowRunner require OAuth?

No. remove.bg connects to FlowRunner with an API key, no OAuth flow required.

Can remove.bg trigger a FlowRunner workflow automatically?

remove.bg doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.

Start building with remove.bg

$100 in credits. No card required. Connect in minutes.

Start Building Free Talk to us