- Creative AI
- Posts
- Midjourney Killer?
Good day, humans.
Table of Contents
Midjourney killer?
Black Forest Labs, the team that helped develop the original Stable Diffusion, has launched Flux, an open-source text-to-image model. Flux is a powerful model that can generate images on par with Midjourney and possibly surpass any other currently available model, whether open or closed source.
One of the problems with this type of open-source application is the difficulty of installation. Unlike Midjourney, Leonardo AI, and similar platforms, where users can just sign up and start using it, Flux needs to be installed on your machine and has high requirements. Fortunately, Flux is available online on Replicate.
Better than Midjourney?
Flux is being considered a Midjourney Killer, the tool that will dethrone the current leader in AI Art generation. Dogan Ural made a quite comprehensive comparison on X.
Why it matters
Flux is an important advancement in open-source models, and while its quality is enviable, it still doesn't reach Midjourney's level. But its progress in such a short time is spectacular. They have already announced that they are working on Video AI as well, so we should stay tuned.
Midjourney has a new version, improving textures and text
Midjourney, the leading artificial intelligence image generation platform, has launched a surprise update to its core model. In version 6.1, human skin looks more natural and rendered text is also more legible.
Runway Gen-3 now has Image to Video
Runway’s Gen-3 has launched the functionality to generate video from an image. Previously, you could only generate video from text. Although the tool's possibilities were already enormous, this new functionality allows for more control over the aesthetics and character design.
Comparison with other AI video tools:
Dream Machine from Luma Labs is a step ahead of the others, as it currently allows you to define the initial and final image of the sequence.
Runway and Kling.ai have the Image to Video feature, and both have a high level of quality in the generated video. However, Runway also has a multitude of additional tools.
At a significant distance, the rest of the AI video tools: Sora is not available, and Pika, Haiper, Krea, and similar tools are still at the previous level of video generation.
Why it matters
AI video tools currently lack the character/aesthetic consistency found in tools like Midjourney, Leonardo, and Ideogram, so their use for longer storytelling is limited. This applies not only to people but also to environments and objects.
Gen-3’s Image to Video feature is a significant advancement that opens up new possibilities in video creation, providing greater visual consistency and allowing for more creative flexibility for users.
White-haired girl in a white hood, in the style of Artgerm, on ArtStation, detailed, wearing a white scarf, with blue eyes and a cold expression, against a snow-white background, white tones, in the style of concept designs by Yoji Shinkawa and Greg Rutkowski, for a game illustration, like a game cover, in a fantasy style.
Canva buys Leonardo AI
Canva has announced its plans to acquire Leonardo.ai, one of the most advanced AI art generation platforms, as part of its goal to build a "world-class suite of visual AI tools."
In Canva’s announcement, it is stated that Leonardo AI will "continue to develop its web platform" as a separate product offering. Leonardo AI’s technology will be "rapidly" integrated into Canva’s existing product suite.
Why it matters
With the integration of Leonardo AI's advanced technology, Canva now has a response to Adobe Firefly and can offer generative AI to its users.
Suno and Udio respond to the legal demand
The AI music startups have responded to the legal demand accusing them of stealing musicians' intellectual property. They have responded that although they trained their models with famous songs, they claim it is legal.
Use this Midjourney prompt
Surreal Fashion --ar 3:4 --v 6.1 --chaos 5 --sref 1000003 --p
Thanks for reading.
See you next week! Hi 👋 I'm Erik Knobl, Product Designer by day and Generative AI explorer on weekends. I share my learnings in this newsletter. Consider subscribing to stay in touch.