Not only math, Gemini 2.5 Pro with #DeepThink also achieved state-of-the-art performances for coding (on LiveCodeBench v6), multimodality (on MMMU), and much more! Check out all other #GoogleIO amazing developments with Gemini 2.5 from @GoogleDeepMind and @Google at
MULTIMODAL AI
-
Veo3 AI Video Generation: Testing Creative Capabilities
By
–
My first two tests with Veo3 were of course "a monkey on rollerskates" and "a wolf howling at the moon." Hard to find much wrong with these! (Sound On)
— Matt Wolfe (@mreflow) 22 mai 2025
I'll post more in the comments. pic.twitter.com/G0oanEyePWMy first two tests with Veo3 were of course "a monkey on rollerskates" and "a wolf howling at the moon." Hard to find much wrong with these! (Sound On) I'll post more in the comments.
-
Deep Think Model Advances Multi-Domain AI Reasoning Capabilities
By
–
Deep Think uses new techniques, allowing the model to consider multiple hypotheses before responding. It's not just about math; it also gets an impressive score on LiveCodeBench , MMMU & more! Learn more:
-
Veo 3 vs Wan 2.1: Physics Performance on Complex Video Prompts
By
–
Tried this prompt with Veo 3:
— fofr (@fofrAI) 21 mai 2025
> televised footage of a cat is doing an acrobatic dive into a swimming pool at the olympics, from a 10m high diving board, flips and spins, there is commentary (not slow motion)
Wan 2.1 still has the best physics for out-of-distribution prompts. https://t.co/SFuzWufSTW pic.twitter.com/S2F1HvtO0hTried this prompt with Veo 3: > televised footage of a cat is doing an acrobatic dive into a swimming pool at the olympics, from a 10m high diving board, flips and spins, there is commentary (not slow motion) Wan 2.1 still has the best physics for out-of-distribution prompts.
-
Imagen-4 Generates Cat Typography Art from Clever Prompt
By
–
Same prompt, Imagen-4 > the word “cat” where the shapes of all the letters combine to form a cat using only clever typography
-
Imagen-4 Long Prompts Generate Superior Comic Illustration Outputs
By
–
I haven't played with Imagen-4 a lot yet, but it seems like long prompts give the best outputs. Here's a Reve prompt on Imagen-4: > A comic book illustration features a character with spiky hair and intense glowing eyes, wearing torn punk-inspired clothing adorned with chains
-
Veo 3 Creates Perfectly Looping Video with Music and Voice
By
–
Veo 3 managed to make a perfectly looping video for this one, music, voice and all.
— fofr (@fofrAI) 21 mai 2025
> a neon splash animation shows "Coming soon on Replicate" and then "Veo 3" against a pink and gold background, funky music, there are people dancing, it's the 80s
We're working on it 🚀 pic.twitter.com/Ub3l9BLMnJVeo 3 managed to make a perfectly looping video for this one, music, voice and all. > a neon splash animation shows "Coming soon on Replicate" and then "Veo 3" against a pink and gold background, funky music, there are people dancing, it's the 80s We're working on it
-
Google Imagen 4 Preview Now Available on Replicate
By
–
Google's Imagen 4 preview is on Replicate https://
replicate.com/google/imagen-4 It's a preview, so things might change. -
Veo3 Released: Google’s Next-Generation Creative AI Model
By
–
Veo2 was already SOTA, but Veo3 is something else! Been having a blast playing with it, and loving all the creativity now that it's out (more in 🧵).
— Oriol Vinyals (@OriolVinyalsML) 21 mai 2025
Me walking to I/O thinking about Gemini3, Veo4, … pic.twitter.com/9w1XwqBAMTVeo2 was already SOTA, but Veo3 is something else! Been having a blast playing with it, and loving all the creativity now that it's out (more in ). Me walking to I/O thinking about Gemini3, Veo4, …
-
AI discovers over 300 hidden Nazca-like geoglyphs in Peru
By
–
¿Y si te digo que una inteligencia artificial acaba de revelar secretos escondidos durante más de mil años?“Un grupo de investigadores japoneses utilizó inteligencia artificial para escanear imágenes aéreas del desierto peruano.
— Juan Merodio (@juanmerodio) 21 mai 2025
¿El resultado? Más de 300 nuevos geoglifos… pic.twitter.com/P3JjjL0BJz¿Y si te digo que una inteligencia artificial acaba de revelar secretos escondidos durante más de mil años?“Un grupo de investigadores japoneses utilizó inteligencia artificial para escanear imágenes aéreas del desierto peruano. ¿El resultado? Más de 300 nuevos geoglifos