Google has unveiled its latest text -to -image model Imagine 4, with the previous version, with a routine promise of “significantly better text rendering” than Imagine 3. The company has also introduced a new delex version called Imagine 4 Ultra, designed to follow more precise text indicators if you are ready for additional payment. Both arrive at the Gemini API preview and Google AI for a limited free test in the AI studio.
Google describes the central imagery 4 model as “going for most of your tasks” which costs per .04 per photo. In the meantime, the Imagine 4 ultra, compared to other image generators such as Dell E and Midgorn, with the promise of “strong” output results “when you need to follow the instructions when you need your pictures.” This model has increased the price by 50 %.
The company showed numerous photos, including a three -panel comedy developed by the Imagin 4 Ultra, showing that a small ship that has been attacked by a large blue … Space lizard? Some sound effects like “Crunch!” And inexperienced, “was !!” This photo immediately followed the bat for the bat and felt fine, not unlike tone rendering from the 3D app.
Read another gesture “In front of Vintage Travel Postcard for Kyoto: Cherry flowers famous pagoda, snow -covered mountains, clean blue sky, dynamic color. “ Imagin 4 output that is on a “T”, though there is a lack of charm in general. Another hiking couple was shown waving above a rock, a fake “event guard” fashion shoot. The images were definitely good quality and followed the text indicators, but they still appear to be highly manufactured.
Imagin 4 is fine and already knows a slight improvement, but I am not obligatory for it-especially against market leaders, Dell-A3 and Madjrani 7. In addition, after the initial rush of enthusiasm, the public is seen getting sick of AI art, which is apparently the main use of the use of spheres on social media as below the articles or articles.
