research @GoogleDeepMind - scaling rl, omni, gemini, world sim, nano banana, veo, imagen, etc…

San Francisco, CA
happy to finally reveal what we’ve been working on, really proud of the entire team for all the hard work 🚀
We’re dropping Gemini Omni: our first step towards a model that can create anything from anything - starting with video. It combines Gemini’s intelligence with our generative media systems - representing a leap forward in world understanding, multimodality, and editing 🧵
5
54
12,084
many such use cases, the way we interact with models will only continue to evolve and become more multimodal imo
Gemini Nano Banana Pro can solve exam questions *in* the exam page image. With doodles, diagrams, all that. ChatGPT thinks these solutions are all correct except Se_2P_2 should be "diselenium diphosphide" and a spelling mistake (should be "thiocyanic acid" not "thoicyanic") :O
21
3,246
glad everyone is enjoying our model, we have no plans on slowing down either 😊
26
3
373
59,083
Robert Riachi retweeted
We're incrementally rolling out Nano Banana Pro to Antigravity. It's not only a step change in image generation but also transforms the entire design process with increased quality.
24
22
516
51,966
remember when we used to look at hands to tell if something was ai generated
At first glance you may think it’s real, but no, it’s Gemini 3.0 Pro Image / Nano Banana Pro. Character consistency is next level 🔥.
11
1,336
some cool examples with Nano Banana Pro 🧵
1
2
18
1,253
Prompt 1: Generate a grid of 12 articles of men's casual clothing that I can wear to work, or out with friends Prompt 2: Now mix and match them to show me combinations of what outfits I can create, use Sergey Brin as the model
1
11
1,027
image quality continues to improve and now in 4k!
1
9
696
Robert Riachi retweeted
Tap, hold, and load in 4k.
10
9
110
22,829
Currently in Japan 🍌 🍌
12
1
158
16,907
Robert Riachi retweeted
🍌🍌
2
2
16
1,775
Exciting last couple of months, we’ll only continue to improve from here 🚀
Introducing Gemini 3 — our most intelligent model that helps you bring any idea to life. Gemini 3 is our next step on the path toward AGI and has: 🧠 State-of-the-art reasoning 🖼️ Deep multimodal understanding 💻 Powerful vibe coding so you can go from prompt to app in one shot 🤝 Improved agentic capabilities, so it can get things done on your behalf, at your direction We're making Gemini 3 Pro available across @GeminiApp, Search (starting with AI Mode), and our developer products, so you can use its incredible reasoning powers in your daily life — to learn, build and plan anything.

ALT A GIF reads “Introducing Gemini 3. Learn anything. Build anything. Plan anything. Our most intelligent model that helps you bring any idea to life.”

13
1,216
Gemini 3 ♊️ ♊️ ♊️
1
9
909
had a 40mn conversation with him and he starts tweeting like this
I’ve never been more confident in Google than I am today
17
9
734
140,843
clearly a bug, will fix in the next version :)
we're so back
45
47
1,994
176,414
just kidding ofc, glad everyone is having fun with the model ❤️
8
161
7,751
some cool examples with Gemini 2.0 native image output 🧵
62
179
3,769
481,792
Gemini can generate pretty consistent gif animations too: 'Create an animation by generating multiple frames, showing a seed growing into a plant and then blooming into a flower, in a pixel art style'
1
2
65
7,910
text rendering is awesome
Native image generation & editing with Gemini 2.0 Flash is out! Check this out -- developers.googleblog.com/en…
2
105
21,107
colorize your old photos
Replying to @OfficialLoganK
It’s soo cool, I’m reviusalising a lot of stuff :)
5
4
183
18,612