Independent Developer Releases Open Source Photoshop Alternative
Independent designer Robbie Tilton released an image editing software named Compositor on September 18, positioning it as a "Photoshop alternative" and choosing to make it completely free and open source. He stated that he initially developed this tool to free himself from Adobe Photoshop's subscription payment model, and later decided to release it publicly.
Compositor is not intended to be a fully-featured Photoshop alternative but is deliberately "narrowly focused," retaining only the core tools that Tilton personally uses in his image compositing workflow, removing what he considers unnecessary features. He candidly mentioned in a tweet that this workflow might seem "a bit outdated" by 2026, but he plans to continue using it until AI tools can achieve pixel-level detail handling with true precision.
Tilton particularly emphasized the significant difference in software size: Compositor's entire application is only 12MB, while the Photoshop installed on his machine is as large as 6455MB, highlighting Compositor's design orientation towards "lightweight".
According to publicly available information on the project’s GitHub, Compositor is developed in Swift, based on native Xcode development, and only supports macOS (minimum system requirement macOS 26). It is released under the MIT open-source license, allowing free use and modification for commercial and private purposes. Its features include layer management and blending modes, non-destructive transformations, selection tools (rectangle/lasso/magic wand), brushes and retouching/cloning tools, adjustment layers, as well as filters like blur, noise reduction, and lens correction, supporting multi-project tab switching. It can import JPEG, PNG, HEIC, TIFF formats and export as JPEG. The release version is code-signed and notarized, distributed in .dmg installation package format.
Notably, according to related reports, Tilton utilized an AI model named "GPT-6 Astra" during the development process to complete the initial functional planning and code structure, followed by manual adjustments and refinements by himself. This development approach reflects the common path of current independent developers using AI-assisted tools to accelerate software prototyping and then refining it manually.
After the project release, the GitHub repository has received 45 stars and 1 fork, with users submitting feature requests in the project issue area for a "Windows version". However, as of now, Tilton has not responded, indicating that the software will remain a macOS-exclusive tool in the short term, not supporting Windows users.
From an industry perspective, small professional software like Compositor, which is "individually developed, custom-made, and open-source released," reflects a persistent long-tail phenomenon in the creative tools sector: facing industry-standard software like Adobe Photoshop, which is feature-rich but has high subscription costs and large sizes, some professional users and independent developers turn to create lightweight alternatives that only cover their core needs, leveraging the open-source community for potential future maintenance and feature expansion. Such projects typically do not pose a substantial commercial threat to industry giants like Adobe but continue to attract a segment of users dissatisfied with subscription models or software bloat.
Source: Public Information
ABAB AI Insight
The practice of independent developers creating lightweight professional tools to escape the constraints of large software subscription models is not new—ranging from the emergence of open-source text editors and code editors (like those escaping commercial IDEs) to similar "escapes" from Adobe's subscription model today, these projects typically follow the same historical path: developers first write code to solve their specific needs, and after validating its usability, they decide whether to open source it and release it to a broader user base.
From a resource investment perspective, Tilton has hardly used traditional capital; instead, he utilized personal time along with AI-assisted tools (reportedly leveraging the "GPT-6 Astra" model for initial functional planning and code structure) to compress development costs. This "individual + AI-assisted" development path is making it possible for a single person to complete prototype software development that previously required a team, further lowering the barrier for independent developers to enter the professional software tool arena.
A comparable historical case is the long-standing phenomenon in the open-source community of "lightweight professional tools challenging commercial giants," such as GIMP versus Photoshop and VS Code versus traditional commercial IDEs—these projects typically do not pursue feature parity but instead focus on being "lightweight, free, and customizable," attracting price-sensitive or non-full-featured niche user groups. The current creative software industry is in a phase where subscription business models and open-source/independent alternatives coexist and evolve, with the proliferation of AI-assisted development tools further lowering the technical barriers for individual developers to launch such alternatives.
This event is essentially a small-scale technological substitution attempt: in the past, creating a professional-grade image editor with layer management, non-destructive editing, and various retouching tools typically required a professional software team to invest significant manpower and time; now, with AI assistance for initial code structure, individual developers can produce a tool that is only a fraction of the size of commercial software but meets their core workflow needs in a shorter time. Mechanically, such projects are unlikely to shake industry giants like Adobe, which have a complete ecosystem and corporate client base in the short term, but they reveal a trend that is happening—AI-assisted development is transferring the ability to create "custom professional software" from enterprise-level teams to individual developers.
ABAB News · Cognitive Laws
Bulky giants will always give rise to lightweight alternatives.
AI does not lower the technical threshold, but the team threshold.
Functionality sufficient for use is the best feature list.