Home
Ai
Ai Features
Top AI announcements from 2025: Nano Banana, Sora, Perplexity’s AI Browser and More

Top AI announcements from 2025: Nano Banana, Sora, Perplexity’s AI Browser and More

The year 2025 was a pretty important one for the AI industry, with new AI models, apps, and tools being introduced.

Written by Akash Dutta, Edited by Ketan Pratap | Updated: 28 December 2025 13:04 IST

Top AI announcements from 2025: Nano Banana, Sora, Perplexity’s AI Browser and More

Photo Credit: Unsplash/Christopher Gower

Meta also created a new AI division, dubbed Superintelligence, in 2025

Click Here to Add Gadgets360 As A Trusted Source

Highlights

Gemini 3 emerged as the most powerful AI model of 2025
Video generation app, Sora, instantly gained popularity among users
Nano Banana was the most viral trend of 2025

The year 2025 was dominated by a handful of artificial intelligence (AI) launches that broke out of research labs and into mainstream conversation. Google pushed its generative AI portfolio into images and language reasoning with Nano Banana, OpenAI turned video generation into a social-first app with Sora, Perplexity expanded its browser ambitions with Comet, and Meta reorganised its long-term AI research under a new Superintelligence division. These announcements stood out not just for what they built, but for how quickly the technology categories themselves shifted in the public eye.

Nano Banana

Google introduced Nano Banana in August 2025 as part of the Gemini 2.5 model family, focusing on AI-powered image generation and editing. The model was released under the name Gemini 2.5 Flash Image, and became colloquially known as Nano Banana due to its viral spread and wide usage in online communities. As per Google, the model is designed to create images from natural language prompts and also edit uploaded images using the same interface.

Google's Nano Banana is integrated directly inside the Gemini app on mobile and desktop, allowing users to generate visuals, modify backgrounds, adjust clothing or objects, and add multilingual text inside images. Google says the image model uses SynthID, a watermarking system that embeds a digital signature into AI-generated images for provenance tracking.

Google, OpenAI and Other AI Companies Sued by Group of Authors

Sora App

OpenAI launched the Sora mobile app in 2025, turning its diffusion-based text-to-video model into a consumer-facing creative tool. Sora lets users input text prompts or reference images and generate short videos with motion and sound. The model can create clips up to 60 seconds long, simulating movement based on the prompt context.

The Sora app rollout included a high-profile, multi-year, $1 billion agreement with The Walt Disney Company, allowing Sora users to include select licensed Disney characters from franchises such as Star Wars and Pixar in generated videos. OpenAI and Disney announced the collaboration in December 2025.

OpenAI said the app is available on iOS and will soon be released on Android, and users can share generated videos directly within supported social feeds, a move that made Sora feel more like a media tool than a developer API demo.

Is Microsoft Really Planning to Rewrite Windows in Rust Using AI?

Perplexity's Comet browser

Perplexity AI announced Comet, its first-of-its-kind AI browser, in 2025. The browser launched on Windows and macOS in July, followed by an Android release in November 2025. As per Perplexity, Comet merges traditional browsing with an AI agent that can summarise webpages, answer questions inside the browsing context, and interpret links or page content without switching tabs.

Perplexity says Comet is built on Chromium, meaning it inherits compatibility with Chrome extensions and the Blink rendering engine. The browser includes Perplexity's conversational AI search assistant, branded separately from Google or OpenAI models.

Gemini 3 AI model

Google launched Gemini 3 in late 2025 as the company's next-generation multimodal language model family, including variants such as Gemini 3 Pro, Gemini 3 Flash, and Gemini 3 DeepThink. Google says Gemini 3 Pro is optimised for complex reasoning tasks across language and images, while Gemini 3 Flash is a lower-latency variant designed for faster responses.

You Can Now Get Google One and Gemini Annual Plan for Half the Price

Gemini 3 models are available via the Gemini API on Google Cloud, and also power parts of the Gemini chatbot experience on mobile devices.

Gemini 3 Flash, released one month later, gained attention for becoming a low-latency feature in Google's AI Mode rollout, where it was surfaced in the browser interface for tasks such as summaries and AI-assisted search.

Meta's Superintelligence Lab

Meta reorganised its long-term AI research efforts under Meta Superintelligence Labs (MSL) in 2025. The division consolidated teams previously spread across Meta AI research groups. As per Meta's announcement, MSL is focused on advanced AI research, infrastructure and large-scale model development, and is led by Alexandr Wang, who previously founded Scale AI.

Meta said the lab is not a product launch division, but a research-first unit dedicated to long-term work on AI systems that could eventually exceed human-level capabilities across domains.

Comments

Get your daily dose of tech news, reviews, and insights, in under 80 characters on Gadgets 360 Turbo. Connect with fellow tech lovers on our Forum. Follow us on X, Facebook, WhatsApp, Threads and Google News for instant updates. Catch all the action on our YouTube channel.

Further reading: AI, Google, OpenAI, Meta, Artificial Intelligence

Akash Dutta Email Akash Dutta

Akash Dutta is a Chief Sub Editor at Gadgets 360. He is particularly interested in the social impact of technological developments and loves reading about emerging fields such as AI, metaverse, and fediverse. In his free time, he can be seen supporting his favourite football club - Chelsea, watching movies and anime, and sharing passionate opinions on food. More