ArtCraft vs Adobe · Seven Free Rust Apps, One Week and a Reality Check

Quick Summary

• Rough Cut Pete tells the story of Brandon Thomas, ArtCraft and seven Adobe-inspired creative applications whose public repositories were created on September 30 and October 1, 2026. The compressed timeline is real, but it describes the first public code and releases, not seven mature replacements reaching parity in a week.
• The original lineup is PhotoCraft, VectorCraft, FilmCraft, LightCraft, PdfCraft, EffectCraft and DesignCraft. By October 12, the official directory had expanded to twelve projects, adding WordCraft, SoundCraft, DeckCraft, CADCraft and GridCraft.
• Each of the seven original repositories is public on GitHub and offers a choice of MIT or Apache 2.0 licensing. The code, notices, bundled assets and ArtCraft trademarks can carry different obligations, so review the exact repository before redistributing a build.
• PhotoCraft is officially described as early alpha. The other major apps are marked early alpha or in development. Their familiar interfaces and rapid progress are impressive, but neither the project nor independent reviewers recommend them as complete daily replacements for professional Adobe workflows yet.
• ArtCraft says the apps are rebuilt in Rust using a clean-room approach, with public specifications and observed behavior rather than proprietary Adobe code. That statement matters, but it does not settle every possible copyright, trademark, trade dress, patent or contract question.
• Adobe remains financially formidable. Its Q3 FY2026 filing reported record revenue of $6.76 billion, $6.58 billion in subscription revenue and more than one billion monthly active users. The immediate competitive question is less whether Adobe disappears and more whether open tools force better price, portability and cancellation experiences.
• The cancellation controversy is not just internet folklore. In March 2026, the US Department of Justice announced a proposed $150 million settlement resolving allegations about early termination fee disclosures and cancellation practices. Adobe denied wrongdoing. The UK Competition and Markets Authority also has an open investigation and states that no conclusion has been reached.
• Treat the Crafting Apps as promising alpha software: download only through the official project or verified GitHub releases, preserve originals, test with disposable copies, verify exported files in a second application and keep client deadlines on proven tools.

Credit, publication date and the short verdict

The embedded investigation was created by Rough Cut Pete and published on October 11, 2026. Pete connects the ArtCraft launch to Adobe's subscription and cancellation controversies, tests the legal and business narratives around it, and ends with the correct practical warning: download from the official source, keep backups and do not trust early alpha software with a client deadline.

The short verdict is that ArtCraft is a remarkable engineering and open-source story, not evidence that seven production-grade Adobe replacements were finished in seven days. The repositories exist, the builds exist and the community response is substantial. The remaining gap is reliability, depth, interoperability and years of edge cases that professional creative software accumulates.

Facts in this guide were checked on October 12, 2026 against the ArtCraft website and repositories, Adobe filings, government sources and an independent hands-on review. Repository statistics and alpha behavior can change quickly, so the linked sources are more useful than any frozen headline.

What 'rebuilt Adobe in a week' actually means

The seven original repositories were created across September 30 and October 1. PhotoCraft's first public release followed on October 2, and wider attention arrived during the following week. That makes the launch tempo genuinely unusual.

It does not mean Photoshop, Illustrator, Premiere Pro, Lightroom, After Effects, InDesign and Acrobat reached complete behavioral parity in seven days. A repository can be created quickly while thousands of commands, file-format cases, plug-ins, codecs, accessibility details, color-management paths and performance problems remain unfinished.

The official PhotoCraft documentation now makes this distinction explicitly. It says much of Photoshop's feature surface exists in some form, but the app is not yet a replacement for daily professional work. It also warns that a wired menu command measures availability, not equivalent behavior.

A fair headline is therefore: a Brandon Thomas-led team and open-source community used AI-assisted development to publish rough, recognizable versions of seven large creative applications at exceptional speed. The accomplishment is still significant after removing the parity claim.

Who built ArtCraft

Pete profiles Brandon Thomas as an Atlanta-based engineer and filmmaker with roughly fifteen years in software and a decade of Rust experience. Thomas previously founded FakeYou and Storyteller, the startup that became ArtCraft.

The project is often described as one developer rebuilding Adobe, but the public repositories credit the ArtCraft team and contributors. Pete also notes a small funded team and a growing volunteer community. Brandon's leadership and AI use are central to the story, but 'solo developer' is now too narrow for the ongoing project.

Claude-assisted commits are visible in the public history, and Pete reports thousands of changes across the launch period. Co-authorship metadata shows that AI participated in the workflow. It does not prove that every generated change was correct, independently designed or production-ready. The real achievement is the human system around generation: architecture, testing, public specifications, review and rapid iteration.

The original seven apps, mapped clearly

PhotoCraft is the image editor. Its stated scope includes layers, masks, adjustments, type, brushes, smart objects and PSD or PSB workflows. It is the most visible project and the one with the clearest early-alpha warning.

VectorCraft covers vector illustration, while FilmCraft targets editing, color and sound on a timeline. LightCraft is a photo library and non-destructive RAW developer. PdfCraft handles reading, organizing, combining, splitting and securing PDF files.

EffectCraft targets motion graphics and visual effects. DesignCraft targets page layout and publishing. These roles roughly correspond to familiar Photoshop, Illustrator, Premiere Pro, Lightroom, Acrobat, After Effects and InDesign workflows, but the projects are independent and are not affiliated with or endorsed by Adobe.

The official directory lists installers for macOS, Windows and Linux. Most app pages also list a Web build. FilmCraft is the important exception in the current directory, which lists desktop installers but not a Web version.

The lineup already grew from seven to twelve

By October 12, the official Crafting Apps directory described twelve projects rather than seven. WordCraft adds word processing, SoundCraft a digital audio workstation, DeckCraft presentations, CADCraft computer-aided design and GridCraft spreadsheets and data analysis.

That expansion supports Pete's larger point about the recipe: AI-assisted teams can scaffold broad application families much faster than before. It also increases the burden of maintenance. Every additional product adds platform builds, file formats, security surfaces, documentation, accessibility, testing and user support.

Breadth is therefore best read as a research and community roadmap. A professional buyer should evaluate the exact task they need today, not assume the presence of an app card means mature parity with a decades-old incumbent.

Free, open source and the licensing details

The seven original repositories each include MIT and Apache 2.0 license files and let users choose either license. Both are permissive open-source licenses. They permit use, modification and redistribution when their conditions are followed.

The practical details still matter. Apache 2.0 includes notice requirements and an express patent license with conditions. MIT requires preservation of its copyright and permission notice. Bundled fonts, icons, images and other third-party assets may retain separate licenses, and the ArtCraft names and logos are trademarks rather than code licensed for unrestricted reuse.

A public repository is also not a warranty. Both licenses disclaim warranties and liability. Anyone shipping a fork, commercial bundle or managed service should inspect LICENSE, NOTICE, ATTRIBUTION and brand files in that specific repository and obtain legal advice for consequential use.

Existing copies received under a valid open-source grant do not simply become proprietary because a future version changes license. That principle does not erase separate trademark, patent, contract or third-party-code questions, and it is not a prediction about any dispute.

Clean-room development and Adobe compatibility

ArtCraft describes the Crafting Apps as clean-room reimplementations written in Rust from public specifications and observed behavior, without proprietary Adobe code, shaders or assets. PhotoCraft says its PSD implementation uses Adobe's published file-format specification and public test corpora.

Adobe does publish a detailed PSD and PSB file-format specification for third parties that need to read and write those files. That supports interoperability work, but a public specification is not a blanket legal safe harbor for every visual, behavioral or branding choice.

Compatibility itself is multidimensional. A document can open while changing type metrics, masks, adjustment behavior, color profiles or smart-object structure. A video can import while losing audio, timecode, effects or media relinking. The only meaningful test is a round trip using the creator's real files, followed by inspection in both the source and target applications.

Is it actually good? What independent testing found

The answer is promising but not ready for general professional replacement. TechRadar tested PhotoCraft and FilmCraft on an older Intel Mac. The reviewer liked the familiar interfaces, fast launch and presence of common controls, but encountered selection problems, limited save behavior in the tested PhotoCraft build, slow FilmCraft operation, no normal edge-drag trimming and unusable audio from the test clips.

Those findings are a snapshot of fast-moving alpha software, not a permanent verdict. PhotoCraft's current repository documents far more formats and a native .pcraft format, which shows how quickly the implementation is changing. It also reinforces why version numbers and reproducible test files matter when comparing reports.

The official project is candid about current gaps, including generative features, missing tools, typography depth, professional workflows and plug-in compatibility. That candor is more useful than a feature-count race. A menu item that exists but behaves differently can be more dangerous than a clearly missing feature because it creates false confidence.

For experimentation, education, lightweight personal work and contributions, the apps are already interesting. For paid production, archival files or regulated work, keep a proven application in the loop until your own acceptance tests pass consistently.

A safe fifteen-minute first test

Download through getartcraft.com or the release page linked from the official Storytold GitHub organization. Avoid third-party mirrors. Where a release publishes SHA-256 checksums, compare the downloaded file before installation.

Create a temporary test folder and copy in one non-sensitive project that represents normal work. Never use the only copy. Record the app version, operating system, hardware and original file hash so the result can be reproduced.

Open the file and inspect the structure before changing anything: layer count, names, fonts, masks, effects, color profile, linked assets, audio tracks, frame rate and duration. Make one visible edit, save to a new filename, close the app and reopen that saved file.

Export the normal delivery format and open it in a second trusted application. For images, inspect dimensions, transparency, color and text. For video, inspect duration, frame cadence, audio channels, sync and random frames near cuts. Stop immediately if the app modifies the source, crashes repeatedly or produces silent corruption.

The migration scorecard professionals should use

Score task completion before visual similarity. Can an editor import a real project, make the required change, save an editable version and deliver the expected output without rebuilding the project elsewhere?

Track fidelity separately: visual match, editable structure, typography, color management, effects, media links, audio, metadata and export conformance. A flattened image that looks right is not a successful migration if the client expects editable layers.

Measure reliability over several files, not one demo. Record crashes, corrupted saves, unexpected changes, memory use, render time and recovery. Include one deliberately difficult file with nested masks, uncommon fonts, large dimensions or mixed media.

Finally, measure ecosystem fit: plug-ins, hardware acceleration, collaboration, review links, automation, accessibility, training and vendor support. The switching cost lives in the whole pipeline, not only the application window.

If the apps are free, how does ArtCraft make money?

Pete correctly separates the free Crafting Apps from ArtCraft's commercial cloud services. The desktop applications can be free and open source while the company sells AI image and video generation, hosted storage or other managed capabilities around them.

That is not a contradiction, but users should keep costs precise. Local editing does not imply unlimited free cloud generation. Review the current ArtCraft pricing and privacy terms before connecting sensitive client assets or enabling paid AI services.

A sustainable business model may actually help the open tools survive. The key questions are whether core local workflows stay usable without a subscription, whether cloud use is clearly optional and whether export and project formats remain portable.

Can Adobe survive? The financial answer is yes

Adobe's Q3 FY2026 results reported record quarterly revenue of $6.76 billion, up 13 percent year over year, subscription revenue of $6.58 billion, net income of $1.83 billion and more than one billion monthly active users across its products.

Those numbers do not guarantee that every product strategy is healthy, but they make an imminent collapse narrative implausible. Adobe has entrenched studio workflows, file archives, training, plug-ins, enterprise contracts, support, review systems and integrations that are difficult to replace at once.

The credible pressure is different. Capable free tools lower the cost of trying an alternative, while AI-assisted development lowers the cost of building one. Adobe must keep earning its place through reliability, interoperability, innovation, fair pricing and customer trust rather than assuming switching costs will do the work forever.

The cancellation-fee dispute, stated accurately

The US government sued Adobe in 2024 over the annual, paid-monthly subscription plan. The complaint alleged that Adobe hid a material early termination fee and created obstacles during cancellation. Those were allegations, not findings quoted as settled fact.

On March 13, 2026, the Department of Justice announced a proposed settlement containing $75 million in civil penalties and $75 million in free services for eligible customers. The order also imposed requirements around disclosures, consent and cancellation. Adobe denied wrongdoing, according to reporting and the company's position described in Pete's video.

The UK Competition and Markets Authority opened a separate consumer-protection investigation into cancellation terms and the fee calculated as 50 percent of remaining annual-plan payments. The CMA case page explicitly says an investigation does not mean the law has been broken and, at the time of this article, lists the case as open.

For users, the operational lesson is simple: identify whether a plan is annual paid monthly, annual prepaid or month-to-month before cancelling. Read the cancellation screen, record the quoted fee and contact the vendor when the amount differs from the contract or local consumer law.

Adobe terms, content access and Firefly

The 2024 terms controversy centered on language allowing automated and manual access to content for operating and improving services. Adobe revised and clarified its terms after creator criticism.

Adobe's current public position is that it does not train Firefly on customer content, does not claim ownership of customer content and trains Firefly on licensed material such as Adobe Stock plus public-domain material where copyright has expired. Its content-analysis FAQ says content is not analyzed to train generative AI unless the user chooses to submit it to Adobe Stock.

Users should still distinguish Firefly training from other forms of service operation, abuse prevention, moderation, product improvement and optional partner-model use. The controlling text is the current contract and the settings for the specific product, account type and region, not a summary on social media.

Could Adobe sue? The answer is not a slogan

A clean-room implementation and interoperability goal can strengthen an open-source project's position, but they do not make litigation impossible. Potential disputes could involve copied code or assets, trademarks, confusing branding, trade dress, patents, contract restrictions or representations made by distributors.

Google v. Oracle is relevant to software interoperability, but it was a specific US Supreme Court fair-use decision about copied Java API declaring code. It is not a universal rule that any compatible software may copy any interface or behavior.

Adobe's publication of the PSD specification supports third-party reading and writing of the format. It does not grant rights to Adobe trademarks, private cloud formats, protected assets or patented techniques beyond whatever the relevant law and licenses allow.

The Yuzu and youtube-dl disputes in Pete's video show that open source can face takedowns, settlements and reinstatement, but their facts are very different from a photo editor. They are context, not reliable predictions. No public Adobe claim against ArtCraft was identified during research for this article. This section is general information, not legal advice.

Four plausible ways the story develops

Scenario one is quiet coexistence. Adobe focuses on customers and ArtCraft keeps improving within a smaller open-source community. This is the least dramatic and often the most commercially rational path.

Scenario two is targeted brand or interface pressure. The project changes names, icons, screenshots, labels or layouts while the underlying code continues. Clear independence notices and disciplined trademark use reduce avoidable confusion.

Scenario three is a broader legal dispute. That would turn on specific claims and evidence, not the general idea of making compatible creative software. It could slow development even without a final court ruling because litigation is expensive.

Scenario four is stronger competition. Adobe improves value, tiers, portability and cancellation, while ArtCraft converts rapid scaffolding into dependable products. Users win most from this outcome because both sides must improve the actual experience.

The real story is the new software-production curve

ArtCraft demonstrates that a small experienced team can use coding agents to create a coherent family of native applications at a pace that would previously have sounded absurd. Architecture, Rust expertise, public formats, shared components and community contributions are as important as the model itself.

AI compresses the cost of a first implementation. It does not compress every later cost by the same amount. Compatibility testing, security, performance, accessibility, localization, documentation, support and trust still require sustained human work.

The durable advantage may belong to teams that publish a narrow working tool, expose its tests and roadmap, accept evidence from users and improve one high-value workflow at a time. A broad launch creates attention. Reliable repetition creates a professional product.

Final verdict · Try it, but keep your parachute

ArtCraft is worth watching and worth testing. PhotoCraft alone shows how quickly an agent-assisted open-source project can move from a provocative idea to a substantial native codebase with real file-format ambitions.

It is not yet a responsible one-click replacement for an established creative stack. The safest approach is parallel adoption: choose one low-risk task, keep the source application available, run a written benchmark and move more work only after repeated success.

Adobe is likely to survive. The more interesting question is whether it can preserve loyalty when alternatives become cheaper to build and easier to try. That competition is healthy, provided creators get honest claims, portable files and clear choices rather than another form of lock-in.