PixelBin is an image transformation platform from Mumbai-based Shopsense Retail Technologies that combines a CDN-backed asset store with URL-driven edits — resize, compress, background removal, upscaling and watermark removal are all triggered by rewriting the image URL, and the same operations are exposed through SDKs for ten languages.
Used this tool? Rate it
Used this tool? Rate it
Pixelbin is an image transformation platform that treats an image URL as the editing interface. Instead of opening a file in an editor, exporting a variant and uploading it somewhere, you store the original once and then request whatever version you need by writing the operation into the delivery address. The official documentation states the definition plainly: "PixelBin is an AI-based image transformation platform that gives you simple tools for storing and transforming images with high quality," and lists resizing, filtering, sharpening, watermark removal, background removal and bulk transformation as the representative jobs it exists to do.
The product is operated by Shopsense Retail Technologies Limited, an Indian company whose Terms of Use names its registered office at 1st Floor, Wework Vijay Diamond, Opp. SBI Branch, Cross Road B, Ajit Nagar, Kondivita, Andheri East, Mumbai 400093. The same entity is named as the developer of the official iOS application, which is a useful independent cross-check: the App Store listing for "Pixelbin: AI Photo Editor" carries the seller name Shopsense Retail Technologies Limited, matching the legal text on the website rather than a marketing alias. That entity also runs a cluster of single-purpose sibling sites — Erase.bg, WatermarkRemover.io, Upscale.media, Shrink.media and ConvertFiles.ai — and this matters more than it might appear, because Pixelbin's AI operations are explicitly powered by those siblings rather than by separately branded in-house models.
There is a visible tension in how the product presents itself, and it is worth naming up front rather than smoothing over. The documentation describes infrastructure: storage, delivery, URL-driven variants. The marketing homepage now leads with "AI Image and Video Generator" and the subheading "What will you make today," positioning the same product as a creative studio for text-to-image and text-to-video work. Both descriptions are currently live and neither is wrong — but a developer arriving from the docs and a creator arriving from the homepage will form quite different expectations of what they are signing up for, and the pricing page is built around the second audience.
The delivery pattern is the product's defining mechanic. A Pixelbin CDN address follows the structure https://cdn.pixelbin.io/v2/{CLOUD_NAME}/{PATTERN}/{FILE_PATH}, where the documentation defines {CLOUD_NAME} as the "Unique ID of an organization … created when a user creates an organization," {PATTERN} as "the transformations that need to be applied to the input file," and {FILE_PATH} as "the relative path of the file from the root folder of the storage along with the extension." A worked example from the docs looks like this:
https://cdn.pixelbin.io/v2/dummy-cloudname/FIDrmb/t.resize(h:200,w:200)/images/transformation/basic-default.jpeg
Individual operations are written as identifier.name(param:value,param:value) — so t.resize(h:200,w:200) names the t identifier, the resize operation and two parameters. Multiple operations chain with a tilde and execute in sequence:
t.resize(h:100,w:100)~t.blur(s:2)~t.toFormat(f:webp)
The practical consequence is that variant management stops being a file management problem. A product listing page requesting a 400px WebP thumbnail and an email template requesting a 200px JPEG are two different strings pointing at one stored original. Nothing is pre-generated, nothing goes stale when the source is replaced, and no build step needs to know about either variant in advance.
The documented basic operations cover the substantive middle of image processing rather than a token handful: Blur, Change DPI, Change Format, Compress, Extend, Extract, Flatten, Flip, Flop, Greyscale, Levels, Median, Merge, Modulate, Negative, Normalize, Resize, Rotate, Sharpen, Tint and Trim. Several are more configurable than their names suggest. Resize accepts width, height, fit, position, background, DPR and an algorithm choice. Merge handles overlays and underlays with mode, position, gravity, height, width, top and left parameters, which is what makes programmatic watermarking and badge overlays feasible. Extract crops by top, left, height and width. Median applies a filter with a size range of 1 to 10, defaulting to 3.
The costing of these operations is the reason the list matters commercially. Official FAQ documentation states that a basic transformation consumes just 0.003 credits, against 1 full credit for an Erase.bg background removal — a factor of well over three hundred. Anyone using Pixelbin primarily as a resize-and-compress delivery layer is operating in a completely different cost regime from someone running AI edits, and conflating the two leads to badly wrong budget forecasts.
The official FAQ is unusually direct about provenance: background removal is "by Erase.bg," resolution enhancement is "by Upscale.media," and watermark removal is "by WatermarkRemover.io." These are sibling products of the same parent company, surfaced inside Pixelbin as transformation operations. For a buyer this is genuinely useful transparency — you can evaluate each capability on its own standalone site before committing to the platform — but it also means the AI quality ceiling is set by those separate products rather than by anything Pixelbin does independently.
On top of that layer, the API page shows Pixelbin has become a gateway to third-party generative models. The text-to-image side lists Flux 2 Pro, Imagen 4 variants and Seedream among others; the video side lists Kling, Seedance, LTX-2, Sora 2 and Veo 3.1. This is a genuinely different product proposition from image transformation, and it is what the current subscription pricing is denominated in.
Less prominent but distinctly more interesting for automation work, the Prediction API set includes NSFW Detection, Detect Number Plate, Object Detection, Object Counter, background detection, product visibility checks and centering detection, plus QR code generation and scanning, OCR text extraction and product tagging. These are not editing operations at all — they return judgements about an image rather than a modified image. A marketplace can screen seller uploads for policy violations, verify a product is actually visible and centred in its own listing photo, or extract text from a scanned document, all through the same credential and billing surface as the resizing pipeline.
Underneath the transformation layer sits a digital asset management system with folder and file organisation, URL-based sharing, and search by file type (JPEG, PNG, WEBP), date and size. Bulk work is handled through presets and Jobs, which apply a saved transformation chain across an entire folder rather than one asset at a time. The Enterprise page adds that organisations can connect existing S3, GCS or Azure storage directly, with no data migration required — meaningful for teams whose asset library already lives somewhere and who want transformation without a relocation project.
Official SDKs span ten languages across two layers. Frontend: @pixelbin/core and @pixelbin/react on npm, Kotlin on GitHub, Swift via CocoaPods, Dart on pub.dev. Backend: @pixelbin/admin on npm, pixelbin on PyPI, Go on pkg.go.dev, C# on NuGet, PHP on Packagist. Checking the registries directly confirms these exist and reports their maintenance cadence: @pixelbin/core sits at version 7.0.0 last published on 22 October 2024, @pixelbin/admin at 4.2.0 from 29 August 2025, @pixelbin/react at 3.0.0 from 16 July 2025, and the PyPI package at 4.3.0. Beyond code, there are Photoshop and Figma plugins, and the platform ships as an official extension inside Fynd Platform, where merchants paste an API token generated in their Pixelbin account to enable prompt-based background generation, focus selection between product and background, and element removal.
High-volume ecommerce catalogues. This is the origin story and still the best fit. A catalogue with tens of thousands of SKUs needs each product photo in a grid thumbnail, a listing hero, a zoom view, a mobile variant and a marketplace-specific aspect ratio. Pre-rendering that matrix is a storage and pipeline burden; generating it from URL parameters against one stored original removes the matrix entirely. Background removal and prompt-driven background generation then handle the visual consistency problem that makes catalogues from multiple suppliers look incoherent.
Frontend performance work. Format conversion, compression and DPR-aware resizing are the standard levers for image-weight reduction, and doing them at the delivery edge rather than at build time means a responsive srcset becomes a set of generated strings rather than a set of committed files. The company's own headline metric — 11 TB+ of bandwidth saved — is framed around exactly this, though it is a vendor self-report with no published methodology.
Marketplace and UGC moderation. NSFW detection, object detection and product visibility checks let a platform enforce listing standards automatically at upload time. Number plate detection is a specific and unusual inclusion that makes sense for the automobile marketplace vertical the company sells into.
Real estate and automotive listings. Both verticals have dedicated solution pages, and both share a profile: photos shot by non-professionals in inconsistent conditions, needing normalisation, enhancement and consistent cropping at volume before they hit a listing page.
Design workflows that need a bridge. The Photoshop and Figma plugins matter for teams where the creative work happens in design tools but the output has to land in a managed, transformable asset library rather than a shared drive.
Document and archive processing. OCR text extraction, DPI adjustment and the restoration set — colourise, denoise, restore — cover scanned material and photo archives. Note the caveat under Limitations: user reports indicate multi-page batch work degrades in ways single-image tests do not reveal.
1. Create an organisation. Signing up creates the organisation whose unique ID becomes the {CLOUD_NAME} segment of every delivery URL you subsequently build. New accounts receive 10 free credits to explore with.
2. Upload assets, or connect storage you already have. Files can be uploaded into a folder structure in Pixelbin's own DAM, or — on the Enterprise track — an existing S3, GCS or Azure bucket can be connected directly without migrating data.
3. Build a transformation in the Playground. The docs point to a Playground for composing chains interactively. This is the right place to learn the syntax, because seeing t.resize(h:100,w:100)~t.blur(s:2)~t.toFormat(f:webp) render live makes the sequencing behaviour obvious in a way that reading parameter tables does not.
4. Save recurring chains as presets. Once a chain is settled — say, the exact resize, compress and format conversion your product grid needs — save it as a preset so the delivery URL references a named pattern rather than a long inline string. This also means changing the recipe later does not require touching every template that uses it.
5. Run bulk work through Jobs. For applying a preset across an existing folder rather than on request, Jobs process the batch server-side.
6. Wire it into your application. Install the SDK for your stack and generate URLs programmatically rather than by string concatenation. For an ecommerce team on Fynd Platform, the path is shorter: generate an API token in the Pixelbin account and paste it into the extension settings.
7. Watch credit consumption before you scale. Because basic transformations cost 0.003 credits and AI operations cost 1 or more, the cost profile of a workload is determined almost entirely by its AI-to-basic ratio. Measure that ratio on a small batch first.
Separate your basic and AI pipelines mentally, and in your budget. The three-hundred-fold cost difference between a basic transformation and an Erase.bg call means a single architecture diagram hides two completely different economic systems. Delivery resizing is effectively free at almost any scale; AI editing is a metered, per-image expense. Teams that model them as one line item are the ones who get surprised.
Test AI operations on a representative sample, not a hero image. The most specific and most repeated user complaint on Trustpilot concerns exactly this gap. One reviewer described de-watermarking page 1 of a 23-page PDF successfully as a free trial, purchasing credits on that basis, and then finding "pages 2 through 23 came out extremely distorted with artifacts added." Free-tier evaluation should mirror the hardest inputs in your real workload, not the easiest.
Assume retries are billable and budget for them. Multiple independent reviewers report that failed or unsatisfactory generations still consume credits. One wrote that a single correction "took 12 generations of the AI model to massage it correctly, which, of course took 12 of my credits, even though I only saved the project once." Another estimated that "about 50% of the time it consumes credits without doing anything." Whatever your credit estimate is, the working assumption should be that the effective cost per usable output is a multiple of the nominal cost per generation.
Never let credits accumulate. Official FAQ policy is explicit that unused monthly credits expire: "If you do not consume all the monthly credits you get in your subscription plan, you loose all the remaining credits." Cancellation forfeits any balance outright, while upgrades and downgrades carry the balance forward. There is no scenario in which holding credits back for later is the right move.
Set a calendar reminder before renewal. The Terms of Use state that fees paid are "non-cancellable and non-refundable, except where expressly required under applicable law," and reviewer accounts of unused annual renewals being refused refunds are consistent with that policy being applied as written. This is a policy to plan around, not to appeal after the fact.
Use presets, not inline strings, in production templates. Inline transformation chains scattered across templates become an unmaintainable dependency the first time a design change requires adjusting them all. A named preset is one edit.
Verify your rights before removing watermarks. Pixelbin's own guidance on this is appropriately blunt: "It depends on who owns the image. If you have the rights or permission to edit it, removing a watermark is generally allowed; otherwise, it may be illegal." The tool performs no rights check of its own — that judgement is entirely yours, and so is the liability.
Pin your SDK version and read the changelog. Given that @pixelbin/core has not seen a release since October 2024 while the product surface has expanded considerably, do not assume the frontend SDK exposes every current capability. For newer generative endpoints, calling the API directly may be the only route.
Ecommerce and marketplace engineering teams get the strongest fit. Large catalogues, multi-channel variant requirements, supplier-inconsistent source photography and a need for automated policy screening all map directly onto what this platform does. The Fynd Platform extension makes the integration path shorter still for merchants already in that ecosystem.
Frontend and platform developers who want image optimisation as a delivery-layer concern rather than a build-time one will find the URL model natural, and the ten-language SDK coverage means the integration language is rarely a blocker.
Marketing and creative teams inside larger organisations can use the DAM, presets and Photoshop/Figma plugins to get consistent output without each person running their own local export workflow.
Automation and operations engineers are served by the detection endpoints more than the editing ones. Counting objects, verifying product visibility, reading text from images and screening for NSFW content are pipeline primitives, not design tasks.
Individual creators and occasional users should think harder before subscribing. The credit model is unforgiving of light and irregular usage — unused credits expire monthly, cancellation forfeits the balance, and refunds are contractually excluded. For someone who needs a handful of background removals a month, a per-use tool or a plan that rolls credits over is a structurally better fit regardless of how the feature lists compare.
Anyone needing an audited compliance posture should ask for documentation before committing. The Enterprise page asserts SOC 2 Type II certification, GDPR compliance, ISO 27001 and a 99.9% uptime SLA. Those are vendor statements; no independent audit report or trust portal is published on the public site, so request the artefacts through sales rather than treating the marketing page as evidence.
Pixelbin runs primarily in the browser, with the full workspace, DAM and Playground available there. The delivery layer is the CDN at cdn.pixelbin.io, which is what any application actually integrates against.
The official iOS app, "Pixelbin: AI Photo Editor" (bundle ID io.pixelbin.app), was released on 16 April 2025 and is at version 1.2.6 as of 10 July 2026. It is a free download, rated 4+ for content, and requires iOS 15.0 or later. Its App Store rating is 3.7 out of 5 from just 19 ratings — a sample too small to support any quality conclusion, and reported here only as a factual attribute of the listing.
On Android, "PixelBin AI Photo Editor" carries a 4.4 rating from 531 reviews with a downloads range of 10K+, an Everyone content rating, and a last update of 9 August 2026. The Play listing's Data Safety section declares no data shared with third parties and no data collected. The store description also carries an explicit content policy: "PixelBin prohibits the generation of sexual, nude, or otherwise exploitative content. All AI-generated content can be reported directly in-app, and violations of our content guidelines may result in account suspension."
Beyond first-party apps, integration surfaces include the Photoshop plugin, the Figma plugin, the Fynd Platform extension, and direct connections to S3, GCS or Azure storage on Enterprise plans.
New accounts receive 10 free credits on signup. The standalone watermark removal tool is more generous at the entry point, allowing three images to be processed without any sign-up at all, plus three further free credits after account creation. Accepted inputs there are webp, jpg, jpeg and png up to 10MB, with a maximum resolution of 5,000 × 5,000 px.
The pricing page currently shows three subscription tiers billed monthly: Creator at $15.00 for 250 credits/month, Lite at $25.00 for 600 credits/month, and Pro at $60.00 for 1500 credits/month. All three are described as including unlimited access to all image and video models and all editing tools — the differentiator is credit volume, not feature gating. The page also offers One Time, Annual and Enterprise billing modes; the annual toggle is labelled 79% off but the page does not render specific annual figures, so check the live page rather than relying on any number quoted elsewhere. Enterprise pricing is not published and is described only as scaling with volume rather than team size.
A credit is the universal unit of consumption, and its value varies by operation. The official FAQ's own examples: one Erase.bg transformation consumes 1 credit; one basic transformation consumes 0.003 credits.
Two policies deserve emphasis because they drive most of the billing complaints discussed below. First, credits do not roll over — unused monthly credits expire, and cancelling the subscription forfeits any remaining balance, though upgrading or downgrading carries it forward. Second, the Terms of Use exclude refunds: fees are "non-cancellable and non-refundable, except where expressly required under applicable law."
One inconsistency between current official documents should be flagged rather than reconciled, because both are live and they do not agree. The FAQ page still explains overage billing using a "Basic plan with 200 monthly credits costing you say $40 per month" with overage "say $0.4 per credit," and separately describes per-GB storage overage charges against a monthly storage quota. No Basic plan and no storage GB allowance appear on the current pricing page, which instead says storage is included at no additional cost and directs users who exhaust credits to buy a one-time pack or upgrade. Treat the pricing page as the current commercial reality and the FAQ's worked examples as legacy illustration — but verify overage terms directly with sales before committing to a volume workload.
Choosing well here depends on which half of Pixelbin you actually need, because the two halves have entirely different competitive sets.
If you need URL-based transformation and delivery infrastructure, the comparison set is image CDN and media API providers. What Pixelbin brings to that comparison is a bundled DAM, an unusually broad detection API set, and AI operations available through the same URL grammar. What it lacks against the established players is independent verification: there is no published third-party benchmarking of its delivery performance, and its documented SDK release cadence is slower than the category norm.
If you need AI image and video generation, the comparison is against dedicated generative platforms. Here Pixelbin's position is that of an aggregator — it routes to Flux, Imagen, Seedream, Kling, Seedance, LTX-2, Sora 2 and Veo 3.1 rather than running its own frontier models. Aggregation is a real convenience when you want one bill and one API for several model families, but it means model quality is not a Pixelbin variable, and pricing comparisons should be made against going directly to the model providers.
If you need one specific operation, going to the sibling product directly is often the more sensible route: Erase.bg for background removal, Upscale.media for enhancement, WatermarkRemover.io for watermarks. Same company, same underlying capability, without a platform subscription wrapped around it.
For an external reference point on how users cluster these products, Trustpilot's "people also looked at" panel on the Pixelbin page lists Pollo AI (4.3 from 5K reviews), OpenArt (4.0 from 1K), Picsart (4.4 from 1K) and Pixella.ai (3.8 from 5K), alongside the sibling sites watermarkremover.io (3.2 from 8 reviews) and upscale.media (2.7 from 4 reviews). The two sibling figures rest on single-digit sample sizes and should be read as directional noise, not as evidence about those products.
Review sentiment diverges sharply by channel, and both readings are real. On Trustpilot the score is 1.5 out of 5 across 51 reviews, with 92% at one star, 4% at five stars, and 27 of those reviews posted in the last twelve months. On Google Play the same product holds 4.4 from 531 reviews. On the App Store it is 3.7 from 19. The Play figure rests on the largest sample and cannot be dismissed; the Trustpilot figure is small but concentrated and its content is specific rather than vague. The most defensible reading is that the mobile app and the paid web platform generate genuinely different experiences, and that Trustpilot is where billing disputes go while app stores are where feature satisfaction goes. On sample-integrity grounds it is worth noting that the Trustpilot profile is claimed (since November 2023), that the platform flags "no recent history of asking for reviews," and that the company "hasn't replied to negative reviews" — meaning there is no vendor review-solicitation bias inflating the score, but the usual self-selection bias toward aggrieved users does apply.
Failed generations consuming credits is the single most repeated complaint. The pattern recurs across independent reviewers over months: retries billed as separate generations, jobs that hang without producing output while still deducting, and one report of the tool returning "some random AI image that I didn't ask for, yet it cost me credits." On Google Play, users report upscale operations failing with "TRANSFORMATION FAILED Software caused connection abort" or processing indefinitely while consuming coins. The vendor's public replies to these reports consistently redirect to email support without addressing the technical cause in the thread.
Billing and refund disputes follow directly from documented policy. Reviewers describe annual renewals charged after a full year of non-use with refunds refused, requiring bank chargebacks; another described support as responsive in timing but whose "only concrete answer I received was they have a non-refund policy." This is consistent with the Terms of Use rather than a departure from them — which is precisely why it should be treated as a structural characteristic of the product to plan around, not an anomaly to appeal.
Batch quality can degrade in ways single-image tests do not surface. The 23-page PDF account, where page 1 was clean and pages 2 through 23 emerged distorted with added artifacts, is the clearest documented instance. It also illustrates why a free single-image trial is a poor predictor of batch results.
No independent media evaluation exists. Searches across IT and startup media returned no editorial review or investigation of Pixelbin as a product. What surfaced instead fell into three excluded categories: the company's own blog and its parent company's documentation; competitor-authored "reviews" on rival vendors' domains, which carry an obvious conflict of interest; and company database entries on aggregator sites, which are records rather than reporting. Absence of coverage is not evidence of poor quality, but it does mean no independent party has verified the platform's performance, uptime or security claims.
Vendor-reported scale figures are unaudited and internally inconsistent with public data. The homepage cites 58M+ total customers, 142M+ total downloads, 5M+ monthly users, 200k+ daily users and 11 TB+ bandwidth saved; the iOS app description references a community of over 150 million creators. None of this is independently verified, and the one public download figure that is platform-reported — Google Play's 10K+ range — sits several orders of magnitude below the self-reported download claim. Different metrics may well be measuring different things across a whole product family, but the gap is large enough that these numbers should not be used in any evaluation.
Compliance certifications are asserted, not evidenced. SOC 2 Type II, GDPR, ISO 27001 and a 99.9% uptime SLA are all claimed on the Enterprise page without accompanying reports or a public trust centre. Request the underlying artefacts before relying on them.
Frontend SDK maintenance has slowed. The registry record shows @pixelbin/core last published on 22 October 2024, well over a year before the current product surface took shape. The backend and React packages are more current, but a frontend integration should not assume parity with the latest platform capabilities.
Watermark removal carries user-borne legal risk. The tool applies no ownership verification, and the official position places the entire judgement on the user: removal is generally allowed if you hold the rights or permission, and "otherwise, it may be illegal."
Content policy is enforced by the platform. The Android listing states that generation of sexual, nude or exploitative content is prohibited, in-app reporting exists, and violations may lead to account suspension — consistent with the 4+ and Everyone age ratings assigned by both app stores.
Watch for impersonation. The company publishes a caution notice warning that fraudsters use fake domains and email addresses "which may appear confusingly similar or identical," along with social accounts using its logo, and states that legitimate communications come only from support@pixelbin.io and that it never requests advance payments or OTPs. Separately, note that pixelbin.com is an entirely unrelated web design studio in Chelmsford, Essex — if you are looking for this platform, the address is pixelbin.io.
There is a free entry point but not a sustained free tier. New accounts get 10 free credits on signup, and the standalone watermark remover allows three images without sign-up plus three more free credits after registering. Beyond that, use is metered against a paid subscription starting at $15.00/month for 250 credits.
Yes. The Terms of Use state that "the Output shall belong to you," and the official FAQ confirms the platform is intended for commercial use across organisations from startups to large enterprises. Two boundaries apply: you grant Pixelbin a licence to use, process, store, reproduce and modify your inputs for the purpose of operating the platform, and you may not resell or commercially exploit the service itself. Note that owning the output does not resolve rights in the input — if you do not hold rights to a source image, editing it does not create them.
Credits are the single unit of consumption, priced by operation complexity: one Erase.bg background removal is 1 credit, one basic transformation is 0.003 credits. They do not roll over. The official FAQ states that unconsumed monthly credits are lost at the end of the billing cycle, and cancelling forfeits any remaining balance. Only upgrades and downgrades carry a balance forward.
Independent user reports across Trustpilot and Google Play consistently indicate yes — retries, unsatisfactory outputs and jobs that hang without producing a result are all reported as billable. The published documentation does not address this case explicitly, so plan on the basis of the observed behaviour: budget for an effective cost per usable output that is a multiple of the nominal per-generation price.
The pattern is https://cdn.pixelbin.io/v2/{CLOUD_NAME}/{PATTERN}/{FILE_PATH}. Operations are written identifier.name(param:value,param:value), for example t.resize(h:200,w:200), and chain with a tilde: t.resize(h:100,w:100)~t.blur(s:2)~t.toFormat(f:webp). Recurring chains can be saved as named presets.
Frontend SDKs cover JavaScript (@pixelbin/core), React (@pixelbin/react), Kotlin, Swift and Dart; backend SDKs cover Node.js (@pixelbin/admin), Python, Go, C# and PHP. There are Photoshop and Figma plugins, an official Fynd Platform extension activated with an API token, and Enterprise-level direct connections to S3, GCS or Azure. Be aware that @pixelbin/core has not had a release since October 2024.
Pixelbin is operated by Shopsense Retail Technologies Limited, registered in Mumbai, India, with the terms governed by Indian law and exclusive jurisdiction in the courts of Mumbai. The same entity is named as the developer of the official mobile apps. pixelbin.com is not related to it — that domain belongs to a web design and digital marketing studio in Chelmsford, Essex, United Kingdom.
The official answer is that it depends on ownership: if you have the rights or permission to edit the image, removal is generally allowed; otherwise it may be illegal. The tool performs no rights verification, so responsibility for that determination — and any consequences — rests entirely with the user.
Both app stores assign general ratings — 4+ on iOS and Everyone on Google Play — and the Android listing states that generation of sexual, nude or otherwise exploitative content is prohibited, with in-app reporting and possible account suspension for violations.
Not publicly. SOC 2 Type II, GDPR compliance, ISO 27001 and a 99.9% uptime SLA are asserted on the Enterprise page, but no audit reports or trust portal are published on the public site. Request the documentation through sales before relying on these claims for a procurement decision.