Image created with Flux Pro v1.1 Ultra. Image prompt: Google, beach towel tableau, cocktail umbrella in clean primary colors beside a lime drink, bright coastal light, photorealistic, editorial, minimal, landscape, vacation, no text overlays
Gemma 3n now available in the Play Store for on-device, internet free with speech, text and image input! Open local AI Assistants are coming to everyone! 🤯 – New on-device speech-to-text and speech-to-translated-text. – Process audio batch inference for clips up to 30 seconds. https://x.com/_philschmid/status/1965742109157188031
Bytedance’s answer to Nano Banana is really good at making high density infographics. “Draw the following system of binary linear equations and the corresponding solution steps on the blackboard: 5x + 2y = 26; 2x -y = 5.” “Create an infographic showing the causes of https://x.com/bilawalsidhu/status/1965838191019307476
google.com/ai https://x.com/OfficialLoganK/status/1964095318254911511
Good news: @NanoBanana is now available in the Gemini API free tier 🍌 for the weekend (under “gemini-2.5-flash-image-preview”). Go build something people want!”” / X https://x.com/OfficialLoganK/status/1964127721711157600
With reinforcement learning, it learns general principles of coordination – allowing it to generate efficient plans for new workflows in seconds. 💡 This research is a key step towards more adaptable manufacturing lines of the future. Find out more ↓ https://x.com/GoogleDeepMind/status/1965040648400351337
Updated & turned my Big LLM Architecture Comparison article into a narrated video lecture. The 11 LLM architectures covered in this video: 1. DeepSeek V3/R1 2. OLMo 2 3. Gemma 3 4. Mistral Small 3.1 5. Llama 4 6. Qwen3 7. SmolLM3 8. Kimi 2 9. GPT-OSS 10. Grok 2.5 11. GLM-4.5 https://x.com/rasbt/status/1965798055141429523
📢 New Model Drop: Seedream 4.0 is live on Yupp! This image model from ByteDance offers text-to-image generation as well as image editing. We dove in with some prompts: https://x.com/yupp_ai/status/1965827081826422990
🚨 ByteDance just released Seedream 4.0 — how does its AI image generation perform? Zhihu contributors share their feedbacks. Let’s have a quick view👇 🎨 Trisimo 崔思莫: ➤ Seed 4.0 vs Nano Banana: Different tech paths. • Nano Banana: multimodal, stronger understanding & https://x.com/ZhihuFrontier/status/1965681077231727069
🚨 New Model Alert! ByteDance’s latest Seedream 4 is ready in the Arena! 🖼️Seedream 4 merges the capabilities of Seedream 3 (Text-to-Image) with SeedEdit 3 (Image Edit). Come and test out your hardest Text-to-Image and Image Edit prompts! https://x.com/arena/status/1965929099370889432
Good news for developers: we now support our SOTA Gemini Embeddings model in the Batch API (50% discount vs regular price) and you can use this model through our OpenAI compat layer https://x.com/OfficialLoganK/status/1965897686558847111
Let’s make embedding models cool again! Well done @GoogleAI @osanseviero @sundarpichai!”” / X https://x.com/ClementDelangue/status/1965784234259628166
We challenged ourselves to build the cleanest, highest-signal factuality benchmark out there. Today, we’re releasing the result: SimpleQA Verified ✅🥇 On this more reliable, 1,000-prompt eval, Gemini 2.5 Pro establishes a new SOTA, outperforming other frontier models. We’re https://x.com/lkshaas/status/1965799946621202719
SimpleQA Verified Benchmark! A New reliable factuality benchmark for measuring knowledge in LLMs from our Research Team @GoogleDeepMind! – Includes exactly 1,000 prompts for evaluating short-form factuality. – Reduced from original 4,326 questions to address incorrect labels. – https://x.com/_philschmid/status/1965806183970652368
My team at @GoogleDeepMind is hiring DevX PM’s, DevX Engineers (fkn DevRel), and more! Pls DM me if you want to work together and are best in class at what you do : )”” / X https://x.com/OfficialLoganK/status/1964842168465686665
We’re open-sourcing everything. All the projects built during the Nano Banana Hackathon are now released as open-source. Since AI Studio supports remixing, you can easily adapt and reuse them. They may not have been huge projects, but it was a great opportunity to truly”” / X https://x.com/arrakis_ai/status/1965001417716072877
Gemini, just like everybody else. From a fascinating blog post about AI agents assigned to play web games, and failing, in large part because vision and computer use tools aren’t good enough: https://x.com/emollick/status/1963968533051617322
Veo 3 and Veo 3 Fast are now generally available in the Gemini API. We’re also announcing ~50% price reduction for both models and support for 9:16 vertical and 1080 HD outputs. Read the blog ↓ https://x.com/googleaidevs/status/1965160822260318702
Veo 3 API updates 🎥 – price drops! Veo 3 is $0.40 / second, a ~46% decrease and Veo 3 fast is $0.15 / second, a ~62% decrease) – support for 1080p videos – support for vertical videos (mobile) Veo 3 is also stable now and ready for scaled production use :)”” / X https://x.com/OfficialLoganK/status/1965193765146296467
Veo 3 is now Generally Available with prices cut by almost 50% and new support for vertical video (9:16). – Veo 3 now $0.40/sec (was $0.75), Fast $0.15/sec (was $0.40). – Generate video with new 1080p HD output. – Create mobile-first video with 9:16 vertical aspect ratio. – https://x.com/_philschmid/status/1965161626761326983
Out of over 2 million open models, EmbeddingGemma is the top trending model on Hugging Face https://x.com/osanseviero/status/1965774422834622648
. @NanoBanana make me a cool banana themed billboard that says “”hello world”””” / X https://x.com/OfficialLoganK/status/1963658089233891664
Goofing around with @NanoBanana in @GoogleAIStudio this morning 😅 https://x.com/OfficialLoganK/status/1964714974452056446
Official @NanoBanana hackathon 🍌 this weekend in SF, $50K in prizes, teams of 1-4, come build with the DeepMind team! Details below 🧵”” / X https://x.com/OfficialLoganK/status/1963756548552888601
🚨New languages coming to AI Mode in Search: Hindi, Indonesian, Japanese, Korean + Brazilian Portuguese. Try it at https://x.com/sundarpichai/status/1965115123330388467
