Год: 2025

Google has introduced Nano Banana Pro, a new image generation and editing model built on Gemini 3 Pro. The company said the model brings reasoning, real-world knowledge and more accurate visual output, expanding on the earlier Nano Banana model released a few months ago.

According to Google, Nano Banana Pro can help users generate visuals from ideas, prototypes, notes and real-time information. The model can also access Google Search’s knowledge base.

“Nano Banana Pro doesn’t just create images. It helps you create helpful content,” the company said, adding that users can produce infographics, diagrams, recipes or snapshots using grounded information.

The company claims the model can render text inside images with greater accuracy and legibility across multiple languages. This includes longer text, stylised fonts, mockups and localised content.

The company also emphasised improved consistency when blending multiple elements. Google says the model can combine up to 14 images and maintain the likeness of up to five people.

The upgraded system introduces new controls for creators, including localised editing, camera angle adjustments, lighting changes and depth-of-field modification. Users can also export creations in multiple aspect ratios and resolutions, including 2K and 4K.

Google is rolling out Nano Banana Pro across its consumer and professional products. In the Gemini app, the model appears under the ‘Thinking’ option within image creation. Free-tier users will get limited access before reverting to the original Nano Banana. AI Plus, Pro and Ultra subscribers will receive higher quotas. In Search’s AI Mode, the model is available in the US for Google AI Pro and Ultra users.

For professionals, the model will be integrated into Google Ads, Workspace tools such as Slides and Vids, and Flow for filmmaking. Developers can access it through the Gemini API, Google AI Studio, Antigravity and Vertex AI.

Google also announced new ways to verify AI-generated content. All images produced by Google tools will continue to include SynthID watermarking. Users can now upload an image in the Gemini app and “ask if it was generated by Google AI,” based on SynthID signals. Free and Pro-tier images will also include a visible Gemini watermark, which will be removed for Ultra subscribers and Google AI Studio developers.

The company said the goal is to support transparency. “We believe it’s critical to know when an image is AI-generated,” Google said.

The post Google Unveils Nano Banana Pro as Its Next Major Image Generation Model appeared first on Analytics India Magazine.

LTIMindtree has announced the expansion of its global collaboration with Microsoft to accelerate Microsoft Azure adoption and drive AI-powered business transformation for enterprises.

As a part of the collaboration, LTIMindtree will enable faster cloud adoption and unlock enhanced business value for joint customers through advanced AI solutions.

As a Global System Integrator (GSI) partner for Microsoft, LTIMindtree is deepening its commitment to enable global enterprises to maximise their cloud investments, Azure commits and achieve faster time-to-value, the company said.

This collaboration underscores LTIMindtree’s ambition to deliver significant growth in Azure-related engagements, leveraging the strength of its 360 degree relationship with Microsoft across all the solution areas.

It combines LTIMindtree’s industry expertise with Microsoft’s advanced AI capabilities, including Azure OpenAI in Microsoft Foundry, Microsoft 365 Copilot, and Fabric.

Additionally, it will enable intelligent decision-making and automation across sectors, deliver secure and scalable cloud modernisation through Azure migration programs, and accelerate Copilot adoption to boost workplace productivity and enhance customer engagement.

As a strategic partner, LTIMindtree has deployed the full Microsoft Security stack, Defender XDR, Sentinel, Intune, Windows Autopatch, and Entra ID, across multiple endpoints, ingesting comprehensive security data monthly for automated threat response.

The security-first approach positions LTIMindtree as a model for secure hybrid and multi-cloud environments.

Complementing this, LTIMindtree claimed that it is leading the way in enterprise AI with internal adoption of Microsoft 365 Copilot.

Guided by a governance-first rollout, Copilot is now embedded across workflows to enhance productivity and accelerate decision-making.

“By accelerating Azure adoption and embedding AI into every business process, we are helping customers move from pilots to productivity, unlocking innovation, resilience, and growth at scale,” said Venu Lambu, CEO & managing director, LTIMindtree in a press statement.

Stephen Boyle, vice president, global system integrators and advisory partners at Microsoft, said that by adopting Microsoft AI across its own enterprise and applying those insights to help customers modernise and scale responsibly, LTIMindtree is setting the standard for transformation.

The post LTIMindtree Expands Partnership with Microsoft to Accelerate Microsoft Azure Adoption, Drive AI-Powered Transformation appeared first on Analytics India Magazine.

Abhishek Singh, MeitY additional secretary and CEO of IndiaAI Mission, has warned that India’s tech and IT services industry could face serious trouble if engineering talent is not upgraded fast for the age of artificial intelligence.

Speaking at the Bengaluru Tech Summit on November 19, Singh said the rise of AI coding tools poses a direct threat to the country’s long-held advantage in software services.

“The ability to use brain power of Indians to solve world problems that has led to the boom in the IT industry… is facing a challenge from… AI code generators and the AI tools that OpenAI, Anthropic, and Google are making. If we don’t turbocharge our engineers with those AI skills, we run a huge risk and we’ll have a lot to lose,” he said.

He pointed out that India built its reputation as the tech garage of the world, yet the skills needed today are shifting fast. AI, data science, and advanced computing have become essential for the next leap in global technology. A slow response from companies could leave them exposed.

The IndiaAI Mission has started fellowships for students working on AI across fields like engineering, medicine, law, and liberal arts. Singh said data labs are coming up in partnership with states and industry to train data annotators, data analysts, and data scientists in tier 2 cities.

MeitY under the IndiaAI Mission, has also launched ‘YUVA AI for ALL’, a first-of-its-kind free course that introduces the world of AI to all Indians, especially the youth.

The mission is also creating tools focused on AI safety. These include systems for bias testing, ethical certification, deepfake detection, and stress testing, which will sit inside the AIKosh platform.

Singh said India’s long term tech strength depends on how fast companies lift the capability of their engineering teams. If the upskilling does not happen at speed, the country could lose ground despite new investments in compute infrastructure and model development.

India’s own push to build large language models is also moving ahead. Singh said Bengaluru based Sarvam is closing in on the launch of its foundation model. Sarvam is one of twelve foundation model projects supported under the IndiaAI Mission, where the government pays for all compute needed.

The post Indian IT Faces Threat From AI Coding Tools from OpenAI & Google, Says IndiaAI CEO appeared first on Analytics India Magazine.

When three entrepreneurs building India’s own foundation models under the IndiaAI Mission came together for a panel discussion, the venue at the Bengaluru Tech Summit 2025 had to be obviously packed. The audience was eager to listen to what they had to say about the country’s AI ecosystem.

Moderated by Kalika Bali from Microsoft Research, the session included Vivek Raghavan from Sarvam AI, Sashikumar Ganesan of Zenteiq, and Ananth Nagaraj of gnani.ai. They discussed why India cannot depend on global systems and must build its own base models, and what this means for the next 800 million Indians who don’t live inside Bengaluru’s ring roads.

The three startups, among the 12 empanelled under the IndiaAI Mission, have all taken up different verticals for building foundational models, ranging from a linguistic focused model to voice and material discovery models.

Raghavan of Sarvam AI arrived at the point directly. “If we don’t put an effort in foundational models, we will become a digital colony,” he said. He explained the risks of depending on open models whose origins and training data are unclear.

He warned that even an open model “can actually be poisoned with a very small amount of data,” and pointed out that many top-performing open models today come from China, like DeepSeek. The implication was obvious. For a country the size of India, relying blindly on such systems made no sense.

He also mentioned figures to back his argument. “Certain early open models had maybe sub 1% Indian data,” he said. Sarvam’s upcoming model, built under the IndiaAI Mission, will be a 120 billion parameter system with 17 trillion tokens, around 15-20% from Indian sources.

Raghavan said that distillation and domain-specific SLMs will become normal as applications scale for agentic AI.

The Missing Link

Sashikumar Ganesan of Zenteiq widened the discussion with his calm approach. “If you do not know how to build the foundational models, then definitely in the next wave we will be behind,” he said. According to him, India missed the supercomputing wave and paid the price. Missing the foundation model wave would be another generational loss.

ZenteiQ.ai, formerly Zentech AI, is building BrahmAI, a scientific foundational model for engineering intelligence, scientific computing, and industrial innovation. Its approach differs from typical LLMs: the model will understand and validate physics-based scientific questions.

“Our focus is Industry 5.0—applications in aerospace, automotive, EVs, energy, and pharma. India relies on foreign software for industrial R&D. BrahmAI is about sovereign scientific AI,” Ganesan had earlier told AIM.

Initial phases aim for 35 billion parameters, eventually scaling to 80 billion depending on compute availability. ZenteiQ has been allotted 2,128 H200 GPUs for the first year, but wishes to scale it further after that.

Ananth Nagaraj of gnani.ai pulled the conversation away from the labs and into a village 200 kilometers from Bangalore. “AI is a necessity for the next 800 million people,” he said. The internet in his village is still used for WhatsApp and YouTube, but the real potential is elsewhere.

“Unless we control the whole tech stack,” he said, “we are prone to all sorts of attacks.” His concern was beyond inclusion, about security, and sovereignty in the most literal sense. And it was about building tech that solves problems for people who don’t have the luxury of English, typing, or even silence around them.

Gnani.ai is developing a 14-billion-parameter multilingual voice AI model with 1.3 crore GPU hours.

Read: India’s AI Push Might Be Pointless Without National Language Standardisation

India’s Unique Problems

Nagaraj explained the everyday problems with a clarity that cut through the noise. Indian conversations are noisy, code-switched, full of background chaos. People talk from speakerphones in buses, fields and markets. “We handle close to 10 crore voice calls. One lakh audio calls every second,” he said.

They need to give a response in 150 milliseconds. It is a different universe. Western benchmarks simply don’t apply.

This is why Indic foundation models must be voice-first and robust enough for railway stations, farms and government offices. Bali reminded the audience of a previous ASR deployment that collapsed instantly because the model had never seen noise like Indian railway stations. Ananth nodded. This wasn’t a theory. This was daily reality.

Ganesan broke down how scientific foundation models work and why they can’t rely on standard transformers. “You cannot mix match. The next operator is not a probabilistic operator,” he said. Physical laws matter. Equations matter. Encoders must understand scientific structure, not only tokens.

India needs such models for materials, energy, climate and manufacturing. He said India’s biggest challenges — energy, EVs, climate threats, materials — won’t be solved by generic chatbots. They need scientific reasoning, not autocomplete.

Raghavan also added that use cases matter, but only platforms change the country. India Stack succeeded not because it solved one problem, but because it created rails for the future. AI, if built right, can play the same role. “AI is an accelerant,” he said.

He warned about the AI divide being even worse than the digital divide.

The only way to avoid that future is to make sure every citizen gets access to it. Not a few thousand engineers. Not a few million users. Everyone.

The post India’s AI Mission Isn’t About Chatbots. It’s About 800 Million People. appeared first on Analytics India Magazine.

Luma AI, a leading frontier AI company focused on multimodal general intelligence (AGI), has raised $900 million in a Series C funding round led by HUMAIN, a public investment fund (PIF) that delivers global full-stack AI solutions, according to the release. AMD Ventures, Andreessen Horowitz, Amplify Partners, and Matrix Partners also participated in the round.

The funding will support Luma AI’s partnership with HUMAIN to become a customer of Project Halo, a 2-gigawatt AI supercluster in Saudi Arabia, one of the largest compute infrastructure buildouts in the world.

The supercluster will enable Luma AI to train and deploy next-generation AI systems capable of understanding and operating in the physical world, going beyond large language models (LLMs) to learn from video, audio, and language data at an unprecedented scale.

“HUMAIN is the perfect partner for this next stage in Luma AI’s explosive trajectory,” said Amit Jain, CEO and co-founder of Luma AI.

Jain explained that to create AI that can help humanity in the physical world and expand understanding of the universe, it is necessary to build systems that can learn from a quadrillion tokens of information, roughly the collective digital memory of humanity, contained in video, image, audio, and language.

HUMAIN is deploying frontier compute infrastructure at impressive speed, and this is critical to achieving Luma AI’s mission.

To this, Tareq Amin, CEO of HUMAIN, mentioned that “Our investment in Luma AI, combined with HUMAIN’s 2GW supercluster, positions us to train, deploy, and scale multimodal intelligence at a frontier level. This partnership sets a new benchmark for how capital, compute, and capability come together.”

The supercluster will support Luma AI in training peta-scale multimodal data, 1,000 to 10,000 times more information than current frontier LLMs, making AI more applicable for real-world tasks. It will also feature next-generation inference systems capable of serving these models globally in real-time.

Luma AI’s flagship model, Ray3, has already demonstrated the company’s ability to transform foundational research into commercial products, deployed across studios, advertising agencies, and brands, including integration within Adobe’s global products.

With this new round, Luma AI plans to expand into simulation, design, and robotics while maintaining leadership in entertainment and advertising.

The company was also the first to launch models within HUMAIN Create, a regional initiative for building sovereign AI models tailored for the Arabic world.

These models are designed to understand cultural context, visual nuance, and linguistic diversity, enabling creators, enterprises, and governments to adopt AI solutions that reflect their identity, values, and sovereignty.

The post Luma AI Raises $900 Million Series C to Power 2GW AI Supercluster in Saudi Arabia appeared first on Analytics India Magazine.