AI NewsOpera introduces a new web standard to let users deploy AI locally in their browser

Opera introduces a new web standard to let users deploy AI locally in their browser

7:53 PM IST · October 27, 2026

Browser maker Opera has introduced a new web standard that lets users to run AI models directly in its browser, without relying on the cloud. The new standard, WebAI, is designed to work with a variety of AI models. Opera name drops Google’s Gemma — an open-weight model that Google first introduced two years ago — as one example. On Macs, WebAI also supports Apple’s on-device foundation model, the company says. But Opera says WebAI isn’t tied to any particular model and is built to support a wide range of open-weight AI systems. WebAI grew out ofprevious workby Opera to bring locally run models into its browsers. Opera is making the new AI features available as an early preview in its flagship Opera One browser. The company says Opera GX, its browser built for gamers and streamers, and Opera Neon, its browser that uses AI agents to perform tasks, will soon get the same treatment. Users have to choose to opt into the feature, according to the company. Open-weight models offer a number of benefits to users. For one, they are typically free to download and use. Their underlying model weights are also publicly available, which gives developers more freedom to customize how they work. Although unlike fully open-source software, open-weight models still come with some restrictions. There’s the additional advantage of privacy, the company said. Users can tap into AI features without having to send their personal information to a company’s servers or pay for a cloud-based AI subscription. “Developers can add AI features without running AI infrastructure or paying for every request,” the company said in its announcement. “Users don’t need AI subscriptions, the information the local AI handles stays on their own device, and they stay in control of the AI on their computer.” The company said WebAI also allows anyone to generate text, classify content, and understand how pieces of information relate, all directly in the browser. Privacyhas become a growing concernas AI assistants become more powerful and gain access to personal information. Users often surrender vast amounts of personal information with AI chatbots, sometimes without a clear understanding of how that data is stored or used. Locally run models offer an alternative because the processing happens on the user’s own device — and not on a company’s cloud server. That doesn’t eliminate every privacy risk, but it does give users a bit more control. It makes sense then that Opera has made digital privacy one of itsprimary selling pointsover the years. There are trade-offs, however. Running AI models locally puts more demand on a user’s computer, which can affect performance. Opera notes that model performance will depend on the device’s hardware.

Latest AI News

View All News →

In China, the AI Bubble Has Already Burst

While global tech stocks teeter, in China they have already crashed as lofty valuations collapse under worries about AI demand and curbs on Chinese exports.

50 minutes ago

View

Nvidia backs Indian startup helping companies run AI models with fewer GPUs

As the AI boom strains computing capacity, Nvidia has backed Indian startupSimplismartin a $14 million funding round, betting on software that helps companies run AI models more efficiently and, in some cases, with fewer GPUs. Nvidia participated in the all-primary Series B round alongside Accel, Shastra VC, and Micromax, with Dallas Venture Capital leading the investment. The deal values the four-year-old startup at about $100 million, including the new funding. Companies are nowadays spending heavily on GPUs to power AI applications, but running those models reliably and economically at scale remains a challenge. Different workloads demand distinct trade-offs between speed, cost, and performance. Deploying and operating AI models to meet business requirements often also requires complex adjustments. Simplismart says its software automates much of that work, tailoring how models run on existing computing infrastructure instead of requiring customers to manage those configurations themselves. “Nobody actually needs the fastest or the cheapest inference,” Amritanshu Jain (pictured above, left), co-founder and CEO of Simplismart, said in an interview. He added that companies instead need to run AI models at speeds and costs that make sense for their specific applications. Simplismart’s software sits between AI models and the computing infrastructure they run on and adjusts how models are deployed and managed to meet customers’ requirements. It says it can optimize factors such as how quickly a model responds, how many requests it handles, and how much computing power it consumes, helping companies get more out of their existing GPU stack. The Bengaluru-based startup offers companies a way to deploy and run AI models through its own platform or within their existing cloud and on-premises infrastructure. It also provides tools to fine-tune models, automatically scale computing resources as demand changes, and monitor performance. Additionally, Simplismart allows cloud providers build and operate their own AI inference services on GPU infrastructure using its software. Simplismart says it helped AI video-generation platform InVideo cut its GPU serving costs by 56% using its software. The startup also works with AI video companies HeyGen and Higgsfield, as well as enterprise AI startup Ema, among others. Moreover, Indian cloud provider Yotta uses Simplismart’s technology to power the serverless AI inference service within its Shakti Studio platform. Simplismart’s annualized revenue run rate has grown 14-fold over the past six months to an eight-figure sum, Jain told TechCrunch. The figure is based on recent customer usage rather than contracted recurring revenue and excludes revenue passed through to infrastructure providers, the company said. He did not disclose the exact figure. More than half of Simplismart’s revenue comes from the U.S., despite the startup having more customers in India, Jain said. Alongside its investment, Nvidia has been working with Simplismart on technical integrations, including around its Dynamo inference software, Jain told TechCrunch. The collaboration involves Simplismart’s engineers and Nvidia’s solutions architects, while Nvidia has also helped introduce the startup to potential customers and partners, Jain stated. Interestingly, the investment presents an apparent paradox as Nvidia, the dominant supplier of AI chips, is backing software that can help companies run some workloads with fewer GPUs. Jain, however, argued that making AI inference more efficient could ultimately benefit Nvidia by making its hardware more attractive to customers and encouraging wider adoption of AI applications. Simplismart is serving customers in an increasingly crowded market for AI inference software, competing with offerings from Nvidia itself, as well as major cloud providers and hyperscalers including Amazon Web Services, Google Cloud, and Microsoft Azure. A growing number of startups are also developing tools to make AI models faster and cheaper to run as there is growing demand in this segment, driven by rising demand for AI applications and the need to make better use of expensive GPU resources. “Where we tend to win is where people have to be more stringent on their SLAs [service-level agreements] and TCO [total cost of ownership],” Jain said, referring to demanding performance requirements and tight cost constraints. Unlike rivals that offer preset options for faster or cheaper inference, Simplismart aims to find a balance tailored to each customer’s needs, he said. Founded in 2022 by former Oracle engineer Jain and former Google engineer Devansh Ghatak (pictured above, right), Simplismart grew out of the pair’s experience building and deploying machine-learning systems. The two met while studying at BITS Pilani, and the startup now employs 58 people, Jain said. Simplismart plans to use the new funding to expand its presence in the U.S., grow its team, and invest further in developing its inference software. Jain told TechCrunch that the U.S. remains a key market for the startup as it looks to build on its existing customer base there. The funding announcement follows areportby Indian English daily The Economic Times that Simplismart was in talks to raise $40 million to $50 million at a valuation of $300 million to $400 million. Asked about the report, Jain declined to comment, saying the startup was not actively fundraising. He added that the Series B was closed and would not be extended.

50 minutes ago

View

General Compute raises $100 million Series A, countering the neocloud disaster narrative

CoreWeave, a publicly traded neocloud that rents out computing power for AI, has seen its stock price fall to a 12-month low, and Elon Musk is flexing $40 billion in debt for buying computing capacity and predicting that “200+ neoclouds are going to see guppy-level death and consolidation.” It’s not exactly the ideal week to announce a Series A for a neocloud. General Compute co-founder and CEO Finn Puklowski is unfazed. He’s bullish on CoreWeave’s chances, and thinks Musk is just talking his book to drive down the prices of the GPUs he wants to buy. Fears of a chip-led collapse are “massively overstated. I have people calling me up every single week asking me to do TPUs, GPUs, any chip, to get them more compute.” Just as important, he says, Anthropic’s bullish S-1 is “settling fears for today.” General Compute has found a niche building its cloud out of specialized inference chips built by Nvidia rivals. The company struck a $400 million deal to purchase new chips from SambaNova—thefirst debt financingof inference chips—and now has added Positron and Cerebras to its portfolio of chip suppliers. A deal with a large customer worth $100 million in annual recurring revenue has paved the way for the company to raise $100 million in Series A funding led by Blackbird Ventures and Abstract Ventures. Valued at $60 million aftera May seed round, General Compute is now worth $800 million post-money to its investors. The startup’s strategy is to zig where others zag. It focused on chips built for inference while everyone else chased GPUs. And instead of taking on a complex effort to build huge data centers, it uses smaller power sources, rents space in other companies’ data centers, and converts former bitcoin-mining sites into places that run AI models. That approach means General Compute isn’t going to catapult into the top ranks of neoclouds or compete with hyperscalers any time soon. But it’s also not subject to the constraints that come with building data centers from scratch. When General Compute has chips, it will have someplace to plug them in. Puklowski expects the company to deploy 75 megawatts of computing capacity in 2027, and he suspects he will have access to 200 megawatts of power—“more than we can probably fund and get hardware for.” Still, the plan has a limit. Puklowski’s financing work isn’t done—”capital is going to be a full-time job for me, both on debt and equity,” he told TechCrunch. The startup has found key customers in cloud companies that specialize in offering developers high-speed inference; General Compute provides the bare metal, or raw servers with no software layered on top, for them to bring their offerings to market. James Palmer, a principal at Blackbird who co-led the round, said his firm’s confidence in General Compute comes in part from an investment in Baseten, an inference cloud company. “We’ve been proximate to their journey over the last 10 months and observing the wild growth story that [Baseten] has gone on, and the demand for open weight inference, and for companies owning their own intelligence,” Palmer told TechCrunch. When it comes to General Compute, he said, “Cerebras, Positron, SambaNova, all need distribution partners who can take them to scale, who can navigate the complexity of the physical world.” Executives at all three chipmakers sang General Compute’s praises to TechCrunch, each noting the importance of the company’s focus on making the economics of AI cloud compute work. “Everyone in this market is selling speed,” Rodrigo Liang, SambaNova’s CEO and co-founder, said. “The real question is whether you can make money doing it.”

50 minutes ago

View