Edited By
Liam O'Connor

A recent announcement reveals that Flux 3's video preview is now accessible on platforms like Replicate, Krea, and Fal. However, the lack of a video model reference has sparked concerns among the community about censorship and functionality.
The reaction from users following this update has been varied, with many expressing skepticism about the capabilities of Flux 3. Some users fondly reminisced about the days just two weeks prior when the prospect of open weights seemed unlikely. Indeed, one user quipped, "I remember the good old days of 2 weeks ago where we were never getting open weights."
Criticism focused on potential censorship, with several users doubting that the model will be as free as its competitors. "Doubt it will be as uncensored out of the box but could be wrong," one commenter remarked. Others shared concerns regarding performance compared to existing models: "Even if they release it right now, what audience do they hope to capture?"
"They need to release those open weights ASAP, or this whole thing is going to turn into a fiasco," warned another user. This sentiment reflects a growing impatience in the community.
The swift arrival of Minimax- has left many users feeling that Flux 3 is playing catch-up. A commenter pointed out the strategic shift in focus by saying, "Minimaxโs strategy is clear: capture the LoRA creator community If LoRA training remains accessible and affordable, Flux 3 could be left in the deep water."
๐ Users express overwhelming concerns about the potential censorship of Flux 3.
๐ The sudden release of Minimax raises stakes for open weights.
๐ค "All the LoRa's and finetunes are going to be for anyway," cautioned a user, highlighting competitive pressures.
The recent developments with Flux 3 demonstrate critical tensions in the creative AI community, particularly as competition heats up. As users await further news on open weights, many wonder if the anticipated release will live up to expectationsโor if it will simply fall short in the fast-evolving landscape of generative models.
Thereโs a strong chance that the upcoming weeks could reveal more about the capabilities of Flux 3, despite the current skepticism from people. Experts estimate around a 70% likelihood that the team behind Flux 3 will address the community's concerns about censorship, possibly by releasing a more transparent version of the model. Additionally, with the competitive pressure imposed by Minimax-, the urgency to innovate and improve performance might lead to rapid updates or even partnerships within the creative AI space. If they can ease user concerns and highlight clear advantages, the model could begin to regain favor in this crowded market.
Looking back at the tech world, the rivalry between Flux 3 and its competitors mirrors the battle that unfolded between VHS and Betamax in the 1970s. While Betamax initially offered better quality, VHS triumphed largely due to its more flexible licensing arrangements and wider availability. Similarly, Flux 3โs success or failure will depend not solely on technical prowess but on strategic choices in community engagement and marketplace adaptability. This historical lens highlights the nuances of innovation competition, where the best technology doesnโt always win, but rather, the one best aligned with its usersโ needs often prevails.