I am not only disappointed, I am massively angry and feel personally betrayed by Supermachine and the founder. As an early supporter, I purchased the Tier 4 back then on AppSumo in good faith. Before buying, I explicitly asked whether I would benefit from future updates with a high tier, which was clearly answered with “Yes” at the time.

Today, after months of silence and countless requests from the community for new, state-of-the-art image generation models, I log in and am faced with the bitter reality: My 4,000 credits are practically worthless. Supermachine has introduced a new internal currency called "Gems," which now must be purchased exclusively to access the truly new and powerful models.

The long-awaited new models have finally arrived, but for us AppSumo users, they are out of reach. It’s not even a matter of using more credits – no, access is completely blocked. We are forced to spend money again. This is a slap in the face for everyone who supported the project from the very beginning.

This is exactly the kind of sneaky business practice that the AppSumo community always warns about: First lure users in with a “lifetime deal,” and once you have a base, exclude the loyal early customers from all relevant advancements. We were completely screwed over as AppSumo users.

And if the argument comes up that the old models are still available: Let’s be honest, 98% of the available models are hopelessly outdated. The results are at the level of 2022 or maybe early 2023 and utterly unusable for professional applications. Even the integrated Stable Diffusion XL models are long from “state of the art.”

If nothing changes about this situation and our credits remain worthless for the truly new, high-quality models, then this is by far the worst and most disappointing deal I have ever made on AppSumo. It goes against everything that supporting startups should stand for.

My conclusion is clear: I will definitely never buy a tool from this developer again and can only warn everyone else against investing money here. The trust is completely destroyed.

-------------------------- UPDATED response to developer feedback--------------------------
I am correcting my emotional title "Absolute Scam," as "fraud" is legally inaccurate. However, the developer's response reinforces my severe criticism of their approach and product. As a long-time supporter, I have spent months trying in vain to improve the tool through constructive feedback and bug reports. This is not about a legal debate but about the moral responsibility towards the original investors.

1. Core problem: Outdated, unstable base instead of relevant updates
New features mentioned by the developer, such as Faceswap or ControlNet, add no value to the platform as long as the underlying AI models are hopelessly outdated. The developer has been ignoring systematic problems for months:
Massive generation errors: Eternal loading times or complete crashes are daily occurrences.
Poor quality: Even the newest model produces blurry images. My well-founded feedback on this was wrongly dismissed with an incorrect aspect ratio.
Free open-source tools like ComfyUI prove with the same models that it is not the AI but Supermachine’s poor technical implementation that is the problem. Instead of fixing these fundamental flaws, a new paywall was introduced.

2. The "Gem" system: Not a fair model, but a classic bait-and-switch
The claim that buying "gems" is optional is insincere. Since all modern, high-quality models are locked behind it, users are effectively forced to purchase. Reputable providers like Merlin AI show fairness by integrating new features through adjusted credit costs for LTD users, thus respecting the original deal. Supermachine, on the other hand, punishes its most loyal customers and devalues their investment. A fair alternative would have been to increase the credit costs for new models, not a complete exclusion.

3. Lack of transparency and unprofessional business practices
The deal lacked clear rules. The promised "Apprentice Plan" was never defined, and there was no public price overview. This lack of transparency allows the developer to change the rules retroactively and exclude valuable innovations. The approach follows a familiar pattern:
Lock phase: Raise capital with vague lifetime promises.
Stagnation phase: Neglect the core product technically.
Paywall phase: Move relevant innovations behind a paywall and devalue the original deal.

This harsh criticism is not an outburst of anger but the result of months of ignored efforts. I detailed technical problems and made constructive suggestions to the developer in numerous emails—without any consequence. The fundamental flaws remain unresolved. My public review is thus the logical result of a failed dialogue.
Many developers understand that early supporters are partners, not cash cows. Unfortunately, Supermachine is not one of them. I cannot recommend anyone to invest here.