Dimon argued stablecoin issuers paying interest should meet bank standards as talks continue in Washington about the CLARITY Act.
BUSINESS
The 4% rule is done — 5 signs your $1 million retirement portfolio can survive the new withdrawal reality
Here’s what financial planners know but many retirees discover too late.
Target is spending an extra $2 billion to refresh stores and tech to bring shoppers back
Investors cheered Target’s quarterly results and outlook, with the stock surging toward a one-year high on Tuesday.
Two major airlines now ban common tech item
Over the last year, the issue around transporting lithium batteries and the power banks that commonly contain them has once again begun coming up after multiple airlines, the TSA and the entire country of Japan all expanded on existing bans around packing these items in their luggage.The lithium-ion battery that allows for portable charging also contains volatile electrolytes that pose a particular risk of catching fire; the additional restrictions come after a string of in-cabin fires on different airlines.While the wireless headphones that have largely replaced cable versions over the last decade also contain lithium-ion batteries in the charging case, their significantly smaller size has largely kept them under the radar from sweeping bans. But amid a bigger spotlight on airborne fire safety, at least two national airlines have introduced new restrictions on transporting portable headphones including the popular Apple AirPods in checked luggage.”Should only be packed in carry-on baggage”: Airlines increasingly crack down on portable headphonesAir New Zealand became the latest major airline tweak its policy to state that not just batteries and power banks but also wireless headphones should not be packed in checked luggage “under any circumstances.” Taiwanese airline Eva Air made similar changes to its policy that it now displays on its website in November 2025.”Batteries, portable chargers or power banks that are not installed in a device must be packed safely in carry-on,” Air New Zealand writes on its website. “[…] The charges cases for air pods, earbuds and in-ear devices should only be packed in carry-on baggage.”Related: Lufthansa is latest to ban use of item many other airlines allowIncidents involving in-cabin fires from a portable charging case have been significantly rarer than those arising from power banks but their potential has pushed some airlines to issue proactive bans.”Bluetooth earphones and their charging case must be carried in your carry-on baggage, as the charging case contains a lithium battery and functions as a power bank,” Eva Air explains to anyone who may be confused as to why their headphones are treated the same as a power bank.
Air New Zealand is among the major airlines that do not allow wireless headphones to be packed in checked baggage.Air New Zealand
Which other airlines restrict wireless headphones in checked luggageSmaller airlines to introduce similar restrictions include Taiwanese regional carriers UNI Air and Tigerair. Neither airline’s restriction on checked baggage transportation affects travelers’ ability to use wireless headphones to listen to music or watch videos in-flight; they were put in place primarily so that any fires that do break out can be spotted and put out quickly rather than spread in the cargo section of the airplane.More Travel News:Airline to launch unusual new flight to Cayman Islands from the U.S.Iranian strike hits major airport, injuries reportedUnexpected country is most luxurious travel destination for 2026US government issues sudden warning on Switzerland travelNo U.S.-based airline currently has wording that specifically singles out portable headphones although Southwest Airlines became the first to change its policy to explicitly state that any power banks brought inside the cabin need to be clearly visible and kept near the traveler at all times rather than be placed within a bag in the overhead compartment.”Do not pack recalled, damaged, or defective batteries, as they can pose a fire hazard,” the Southwest now also has in its policy on which items can be brought aboard.Related: Airline loses license over false pilot records, all flights off
REI’s waterproof hiking boots can go for ‘hundreds of miles’ — and they’re on sale for $100
TheStreet aims to feature only the best products and services. If you buy something via one of our links, we may earn a commission.Why we love this dealThe end of winter is only a few weeks away, and as temperatures warm up, many people are eager to get outside and enjoy nature again. It’s the perfect time to plant a garden, spend time on your patio, and get active with walks and hikes. But since spring is known for rain, you may want to bring a raincoat — and a sturdy pair of waterproof hiking boots.REI is currently having an impressive sale on the Vasque St. Elias Waterproof Hiking Boots. Normally $240, they’ve been marked down to just $100. Made of full-grain leather and waterproofed with Gore-Tex, these high-quality boots can accompany you on the toughest trails and keep your feet dry.Vasque St. Elias Waterproof Hiking Boots, $100 (was $240) at REI
Courtesy
Why do shoppers love it?Vasque boots have been around since 1964 and have long been a beloved choice for hikers. Sadly, parent company Red Wing Shoes announced in 2024 that it would shutter the brand, saying, “We do not believe that there is a viable path forward.” No new shoes have been manufactured since, which is sad news for fans, but reduced prices like this REI deal mean you have one last chance to pick up a pair before they’re gone for good.The St. Elias boots have a reputation for being a high-quality product. This pair features full-grain leather, which breaks in more easily. They’re also eco-friendly, as the toe boxes and heel counters are made of 60% recycled content, and the laces are made of 80% recycled content. Vasque’s signature Vibram Megagrip outsoles dig into the trail as you hike, proving much-needed traction when the path gets steep. And since these boots are treated with Gore-Tex, no water will get in, which means your feet will stay dry during rainy hikes.Both the Taupe and Cognac colors are on sale, but sizing is limited in the Cognac, as many have sold out, so the Taupe may be your best bet.Details to knowColors: Taupe and Cognac.Material: Full-grain leather with recycled content.Sizes: Women’s 6 through 12, including half sizes. Related: Walmart’s bestselling $160 farmhouse console table is on sale for only $89Shoppers have good things to say about these boots. One shopper wrote, “I have had these shoes for over four years. I took them up Kilimanjaro, and they stayed dry despite rainy days and mud. I have hiked or walked probably hundreds of miles in them, and they have held up great. After all those years and miles, I think the support is starting to wear out, but I plan on buying the same ones again.””Great boots!” a second shopper wrote. “Better fit than expected! They are not too narrow. The sole is sturdy. Love them!”Shop more deals Vasque Horizon Mid Waterproof Hiking Boots, $104 (was $140) at REIVasque Horizon Low Waterproof Hiking Shoes, $$84 (was $130) at REIVasque St. Elias Men’s Waterproof Hiking Boots, $100 (was $240) at REIHigh-end hiking boots can cost hundreds of dollars, but with REI’s deal on the Vasque St. Elias Waterproof Hiking Boots, you can save $140 on a pair of boots from a brand that’s been loved by hikers for decades. Grab a pair now before your size sells out.
38-year-old sporting goods giant closing stores
While Dick’s Sporting Goods has proven resilient and has consistently delivered strong results, many of its rivals have struggled, and a number have gone out of business.Bob’s Stores, Next Adventure, and Moosejaw all shut down in 2025, while Eastern Mountain Sports closed many of its locations. Eddie Bauer is currently in the process of closing its stores in the United States, and Orvis has closed more than 35 locations.Retailers in the outdoors space have been trying to decide how to move forward, Gartner Analyst Kassi Socha told Modern Retail.“A lot of outdoor retailers right now are in planning mode. They’re reevaluating everything from their loyalty programs to their omnichannel experiences to prepare and adjust for the future of what their retail sector looks like,” she said.The 38-year-old Sportsman’s Warehouse appears to be the next sporting goods and outdoor retailer to tighten its belt and close stores.Sportsman’s Warehouse has struggled Sportman’s Warehouse, which operates 146 stores across 32 U.S. states, recently shared a preliminary fourth-quarter and full-year earnings report.Some of the highlights include:Net sales of approximately $334.9 million, and same-store sales of approximately $333.6 million.Adjusted EBITDA of approximately $9.6 million.For the 52 weeks ended Jan. 31, 2026, the company expects to report net sales of approximately $1,209.2 million and same-store sales of approximately $1,205.6 million.This is an expected increase of 1% and 1%, respectively, compared to the prior year.Adjusted EBITDA of approximately $27.5 million.Net debt of approximately $90 million, a decrease of 6.1%.Ending inventory of approximately $312.9 million, a decrease of $29.1 million or 8.5% compared to the prior year.Total liquidity of approximately $107.8 million.In a challenging environment, those numbers should be seen as a positive, but a consensus rating of analysts compiled by TipRanks still rates the stock as a “strong sell.” Fourteen analysts who cover the stock saw it as “bullish,” while five were neutral and three rated it as “bearish.”
Sporting goods stores face a number of headwinds.Shutterstock
Sporting goods and outdoor goods retailers have struggledLike all retailers, sporting goods chains have lost some sales to the internet. Personally, I’ve moved to buying sneakers online since the Covid pandemic, which mirrors the shift many consumers have made toward digital channels.”Mall stores have lost shoppers to stand-alone and online operations in the past two decades, a trend mirrored in the athletic-goods space,” David Swartz, senior equity analyst at Morningstar Research Services, told The Washington Post.That has contributed to a number of closures in the space, a long list that includes the following.Eddie Bauer, filing Chapter 11 and closing North American stores: Outdoor apparel chain Eddie Bauer LLC filed for Chapter 11 bankruptcy in early 2026 and is winding down its physical retail footprint across the U.S. and Canada, with roughly 175-180 store closures planned as part of liquidation or sale efforts, reported TheStreet.Orvis, planned closures of full‑price and outlet stores: Orvis revealed plans in 2025 to shutter 31 full‑price stores and five outlet locations nationwide as part of strategic refocusing due to rising costs and changing consumer demand, according to another TheStreet story.Bob’s Stores & Eastern Mountain Sports: Its parent company filed for bankruptcy and announced the permanent closure and liquidation of all Bob’s Stores and a reduction/closure of many EMS locations in 2024, RetailDive reported.Big 5 Sporting Goods, multiple closures in 2025: Big 5 announced eight store closures in 2025 as part of ongoing footprint optimization while adjusting to market pressures, TheStreet also reported.Next Adventure: The longstanding Portland, Oregon‑area outdoor gear retailer Next Adventure announced plans to close all of its locations by late 2025 amid owner retirement and market challenges, the company shared on its website.REI, select store closures: REI will close specific locations as part of broader store portfolio adjustments in 2025-2026, according to USA Today.All sporting goods chains face new challengesSporting goods chains also face political uncertainty, which could raise costs and lower sales.”The industry has another challenge on the horizon: Eighty-four percent of sporting goods executives expressed concern about the impact of the geopolitical environment on their business. Potential tariff increases this year could have a significant effect on the sporting goods sector, particularly in terms of pricing and supply chain management,” according to McKinsey’s 2025 Sporting Goods Industry report.In addition, sporting goods retailers face another challenge: low physical activity levels among segments of the population.”Physical inactivity also presents an existential risk to the sporting goods industry. If levels continue to rise or even remain constant for younger generations, the market related to physical activity will decline. However, converting physically inactive segments is the biggest potential opportunity for the sporting goods industry,” McKinsey added.The global population that is currently not meeting WHO’s recommended levels of physical activity totals 1.8 billion, an untapped market equivalent to twice the size of India’s adult population.”The good news is that sporting goods companies have an opportunity to take targeted action to empower sedentary consumers to increase their physical activity levels. They could seek to remove barriers to physical activity for more sedentary segments, including via product innovation, marketing campaigns to raise awareness, and enhanced youth engagement,” the study shared. Related: Another alcohol, hard cider brand files Chapter 11 bankruptcy
ONward and upward: An interview with On Holding CEO Martin Hoffmann
On Holding reported earnings Tuesday morning, flaunting record sales and stronger profitability in its final quarter of 2025 — evidence that the company still has pep in its step.Since coming to market during the pandemic, the Swiss footwear company has commanded a premium valuation on the market thanks to its rapid growth. In the fourth quarter, the company’s net sales grew 35.6% on a constant-currency basis, buoyed by a doubling of sales in the Asia-Pacific region, as well as an expansion into apparel and accessories.That growth rate has dwarfed the figures posted by many of the company’s competitors, but on Tuesday, analysts focused on one figure in particular: In its 2026 guidance, On Holding sees its net sales growing just 23%, which some consider to be a sizable deceleration. Some outlets called it weak, and investors hammered the stock in the pre-market before an intraday recovery.But even with that deceleration, there were a bunch of bright spots. To hear more about them, I sat down with On Holding Co-CEO and Acting CFO Martin Hoffmann to talk about the brand, the earnings, and how they’re thinking about new technology and innovation.Note: This interview has been edited slightly for clarity and readabilityStarting off our conversation, I asked Hoffmann to highlight some of the things he wanted to single out from the earnings report:I think summarizing results, clearly we had a standout Q4, very strong momentum really in every region of the world across all the different channels, exceeding expectations. But it’s always important for us. Q4 is not a sales moment; it’s more where you usually harvest all the work that you have done throughout the year, where we continue to win fans at full price and not engage in the discounting that is going on everywhere else.But I think what’s more powerful is looking at the full-year numbers and basically where the Q4 led us to. For the first time, we exceeded the 3-billion mark, which is super, super important. If you convert this to US dollars, that’s almost 4 billion. Not only are the sales numbers exceptional, but also the quality of all the financials. We are at 36% growth, so there is clearly strong momentum.We have gained a D2C share of 110 basis points, so we have grown our D2C channel stronger than our wholesale channel, which is a result of our expansion into our own retail. And this, put together with that vision to be the most premium global brand, is driving a very strong cross-profit margin profile. Despite external forces like tariffs, we have achieved a cross-profit margin that we clearly didn’t believe would be possible two years ago.But 62.8% just shows the power of the brand. And this allowed us to actually reinvest into the product and technology, but also in product building, and at the same time, still driving an EBITDA margin of 18.8% which is also beyond the number that we had originally targeted for 2026. So, it’s a really strong profile, exceeded a billion in cash as well, which I think gives a lot of confidence in dreaming big in the future.And this clearly shows that what we are doing is working and there’s a lot of confidence moving into the future.You guys brought in these results at a time that businesses are traditionally looking to move a lot of inventory through discounts, promotions, Black Friday, and what have you. Year-over-year, your inventory stayed the same while you guys focused on that premium angle. In some markets, like the U.S., there are a lot of worries aboutcomparable brands that are battling with that premium distinction. That’s to say, a lot of your success here has come at a tryingtime for consumer brands. Can we talk about the factors that make you guys different? What’s giving you strength?Is it this pivot from being a “performance brand” focused on footwear to being a “lifestyle brand”catching all sorts of apparel spending?So I think … first, I do not fully agree with the introduction of saying we pivoted from performance to more lifestyle. I think performance is still at the core. What you see in today’s world is really [that] sports is the new fashion or movement is the new luxury, and so products that are deeply rooted in performance but also have the right design language are attractive for a lot of consumers who want to make them part of a longer period of their day.And I think this is where ON is so strong. That performance DNA is just part of our culture from the very beginning. This is how we have built a company out of Switzerland, where you either build a premium or luxury brand, but you never build a mass market brand. And I think this requires a lot of discipline, but it also always needs to be rooted in very strong, innovation-led products.And I think this is where we have our strengths and keep on innovating our products and not compromising on the product. And then if you think about what is premium, then in the end, premium is only something that the customer can feel. It’s this emotion that you have for a premium brand. So when you engage with Apple or Dyson, whatever is a premium brand in your perspective, then you have a certain feeling and certain expectations.And as a brand, you can curate them, and you need to make sure that you are consistent in each and every touchpoint. So when you come into the store, when you come to the website, when you get the parcel … this all needs to be the same [premium] experience you expect. And this is what we spend a lot of time and energy on — just elevating this experience. But all of this is always rooted in strong products that speak [for themselves]. They come from a performance core, but they combine the design language of a fashionable product as well.Over the last few years, you guys have expanded the business into a number of new apparel categories. How do you guys think about investing in those categories and maintaining them, whether that’s R&D, investing in brand, like you just talked about, and the experience? And I think, most importantly, how do you maintain that quality without giving up the quality of the sourcing or the quality or integrity of the underlying product? I think this comes from a clear understanding that … being a premium brand and having pricing power is something you need to earn. It’s not something you can dictate. So you need to earn this through your products. You need to earn this through innovation. It’s the result. It’s not the starting point. And if you don’t get this right, the consumer will see through it very quickly and not accept it.We have more than 100 team members in the innovation department working on innovating on materials, on new foams, and around sustainability. And I’ll just take two examples that will hit basically the shelves and the consumer this year. One is LightSpray, which you probably have seen. We opened our first factory in South Korea, where now 32 robots are basically spraying shoes in a fully automated way.This is not only a manufacturing revolution; it’s also a product evolution. It’s a super light product, but at the same time, because you connect the upper directly to the outsole, the performance is just at a very different level compared to a traditionally manufactured product. So this is the one thing that will come basically to the consumers now. The other one is, we are currently in a phase where there’s a lot of evolution — or almost revolution — happening around foams that we are using on running shoes.And for the first time, we are able to combine this next generation of foams with our CloudTech, so with the holes in the shoes. And this really will elevate the sensation that you have when you use our products in the future. Comes in October. So ,those are two examples where we have the most advanced foam laboratory outside of Asia in our headquarters in Zurich. We built that LightSpray technology in Zurich. So those are examples of like really deep innovation that we’re doing.On that note, I wanted to maybe talk about supply chain modernization and maybe some of the trends you’re taking up while investing to build great products. On the product front, you mentioned robots. Right now, a lot of people are talking about automation and new technology, and I’m wondering where specifically you guys are making a point of investing? I think one of the key goals that we have is to continue to automate manufacturing.Our footwear manufacturing or sports footwear manufacturing is highly manual today. There are almost 150 people involved in creating a product today. So moving away from this and basically achieving a higher degree of automation, I think this is a super important part of our future. This will be the starting point, and to also be able to think about new places to manufacture products. With LightSpray, we have now gone basically into a fully automated world where we can basically produce the products in each and every place in the world because we are far less dependent on local labor expenses.It is something that we will do as part of this shortening of the supply chain. And the lead times are super important in order to be able to react quicker to new innovations and to consumer demands. So by having a more automated manufacturing, by maybe being able to move manufacturing closer to your distribution markets, and then also shortening the timeline in your supply chain … those are very important pieces to be closer to the consumer.I think the last piece I would call out is really the power that we see now coming from AI into the whole area of supply and demand planning, which of course for a premium brand is one of the most important things to do right because you know, an excess of inventory is the most dangerous element to your premium position, because in the end, it will require discounts, and basically, that would spread messagesthat you don’t want.As you were just talking about planning and regionality, I am wondering how you can react to trends and implement them in the product and the vision with this new angle of attack.Yeah, clearly, I think you can just think of the future in a more flexible way than the past, where the local labor expenses were among the most important factors to consider when deciding where you produce. That’s the goal that we have with automation. But then also, automation drives more quality, which I think is an important piece as well. But at the same time, all of this needs to make sense from a consumer perspective. You don’t want to near shore just to near shore. Currently, the quality of product that we get from our factories is outstanding. There’s a lot of know-how in how to produce a product and how to innovate on the product in Asia. This is clearly something we want to tap into. This is why the first factory for light spray is in South Korea, because you have that knowledge there.I think there are many factories. I think being more flexible around geopolitics is important. Yes, tariffs. Clearly, it’s thriving, also funding into that area because it will allow you in the future to maybe produce in your distribution markets, which allows you to not pay the tariffs. So there are many aspects in here, but it’s quite a complex infrastructure in the end.At this point, we decided to jump into the Q4 results and the guidance picture for 2026:Let’s talk a little bit about the outlook. Right off the bat, that net sales figure: 23%, down from 36% last year. Analysts might look at the net sales as decelerating quarter-over-quarter, despite the impressive growth in 2025. Recently, markets have been penalizing any sign of deceleration, especially in consumer brands. I’m wondering if we can touch on the factors feeding into that. Is this market or consumer-specific? What’s contributing to the figure here?I mean, the way that I look at this, this is still an almost unseen growth rate at the size of company that we are. And I think it just shows the power that we have in the brand and the confidence that we have in the brand. And at the same time, yes, we could grow faster, but we put a lot of focus on growing in a premium way, to grow in a highly qualitative way, and to be very selective on the wholesale partners that we work with … to be very selective on the new retail locations that we are opening in. And for us, it’s all about the long term and just being able to build a brand that can maintain very strong growth for a long period of time.It’s something that we have proven. And at the same time, if you look at the profitability, we are well ahead of our original investor plan. We actually guide for a higher cross-profit margin this year.So I think that the numbers are really strong and speak to the confidence that is there. We expect that our D2C channels continue to grow faster than our wholesale channels, which is an important piece to really connect with the customers to make use of the power of AI in the new ways of engaging with customers. And then one category for us that is super exciting is apparel. It’s still a small part of our business. Apparel and accessories get us like 7% of sales, but it’s clearly growing much faster than the footwear business. And of course, this is a new part of our business that we are scaling as a company, and that will be an important part of the growth story going forward.Speaking of sales growth and areas for opportunity, Asia-Pacific was a standout, with sales nearly doubling year over year. I wonder if you can talk a little bit about your region mix and how you guys are thinking about each of these regions, plus how they contribute to the picture for 2026.So as a global company, we really look at the global business, and we are growing the company across all regions. When we speak to premium growth, we clearly look at “what are the regions where we have a strong momentum?” or “where are we accelerating?” and “which regions can we just also take the foot a bit off the gas?” And this is constantly evolving across different regions.In Asia-Pacific, we’re experiencing that that strategy of “premiumness” connects so, so strongly with the consumer and the sentiment of the consumer at that time. If you go into shopping malls in China, Singapore, or Bangkok, the quality and the premium expression that the customer is experiencing there is at a very different level compared to other parts of the world. This is why I think a premium brand like On resonates so strongly with the consumer. And at the same time, we are still at the beginning; we will see the strongest growth coming from the Asia region. Of course, always doubling [sales] is a big aspiration.But then, if you look at Europe, we had done some strategic adjustments two, three years ago; they’re clearly paying off, and you see Europe accelerating on the growth side. And at the same time, also the U.S. will, of course, continue to be, from an absolute perspective, the growth engine.I wanted to dig in a little bit on the D2C/wholesale mix. Obviously, it seems like one big question in all consumer brands right now is what the right mix is. I think the aspiration of most brand-forward companies favors D2C, but you see some brands making bigger investments in wholesale. Obviously, for you guys, it seems to be quite a balancing act.How do you think about this from region to region, or even product to product?There are, of course, differences country by country. I mean, clearly, China is probably the strongest D2C market because you simply don’t have a lot of wholesale partners. So you have a very high DTC share. Whereas you have other markets, like Japan, for example, where you have very strong wholesale partners, you can actually connect this customer very strongly to wholesale. When we think about channels for us, it’s very important to be where the customers are.So we see the strengths and the beauty of being a multi-channel brand, and of really using the different channels to bring the right products to the right customers through the right channel. So this is how we think about this. But then you also have categories like apparel, where more than 60% of the sales are done through our D2C channels compared to the 44% DTC share that we have in the overall business.This is because, when you think about your D2C channels, this is where we can show the full assortment, whereas our wholesale partners always have a selection of our products. Retail stores show most of our products; our e-commerce platform shows our whole assortment. And this is, of course, more important when it comes to how we can really create a brand world and show how we have thought about the products from head to toe or toe to head.Whereas in footwear, if you put a single shoe on a wall, it’s still selling. And this is why. The apparel company that we are building will have a higher D2C share in the future. And this is actually the beauty: The categories that we are entering are increasing our D2C exposure. They drive a superior margin profile into the brand, and they allow us to connect more closely to the customer.So, the evolution of product and the evolution of the premium business model, that really goes hand in hand. And this is what we sometimes describe as this premium growth.With all the talk about AI right now, I am wondering how you guys think about implementing that into your business, whether thinking internally or as part of the customer journey?Yeah, I mean, I think AI is changing every part of our business, but there are three areas where the impact at the moment is the most tangible and where we are also focusing on. One is product creation, so being able to shorten development cycles and also improving the sustainability footprint in product creation, I think is important. The second one we already spoke about is the whole supply chain, and really being able to match demand and supply in a way that we would not be able to do in the past.And then the third one is really elevating the connection that you can have as a customer, and the insights and the engagements that you can have as a customer.We recently brought our new customer service platform to life, which is a conversational platform, but this is not just for talking about warranty cases or return cases. It’s really to engage with the consumer in a conversation and to learn much more about you than what we learned in the past from you clicking certain buttons. I think we will be able to use it to offer you something in the future that is much more specific to you and really allows us to drive new levels of engagement with the customers.I think those are the three key areas that we’re working on.Is there anything else that maybe I didn’t cover that you guys are particularly excited about, whether in this earnings report or plans for 2026? I think you touched on a lot. Of course, behind this is always a team and a culture. And I think that’s something that sometimes gets forgotten by talking about the numbers.But in the end, having people who can be the creatives, but also executors of sending millions of pairs of products around the world. I think bringing those two things together just makes this the superpower of the brand.
Google releases Gemini 3.1 Flash Lite at 1/8th the cost of Pro
Google’s newest AI model is here: Gemini 3.1 Flash-Lite, and the biggest improvements this time around come in cost and speed, especially for enterprises and developers seeking to leverage powerful reasoning and multimodal capabilities from the U.S. search and cloud giant.Positioning it as the most cost-efficient and responsive model in the Gemini 3 series, Google is offering a solution built specifically for intelligence at scale. This launch arrives just weeks after the February debut of its heavy-lifting sibling, Gemini 3.1 Pro, completing a tiered strategy that allows enterprises to scale intelligence across every layer of their infrastructure.Technology: optimized for the “time to first token”In the world of high-throughput AI, the metric that often dictates user experience isn’t just accuracy—it’s latency. For real-time customer support, live content moderation, or instant user interface generation, the “time to first answer token” is the primary indicator of whether an application feels like a tool or a teammate. If a model takes even two seconds to begin its response, the illusion of fluid interaction is broken.Gemini 3.1 Flash-Lite is engineered specifically for this instant feel. According to internal benchmarks and third-party evaluations, Flash-Lite outperforms its predecessor, Gemini 2.5 Flash, with a 2.5X faster time to first token. Furthermore, it boasts a 45 percent increase in overall output speed — 363 tokens per second compared to 249. This speed is achieved through what Koray Kavukcuoglu, VP of Research at Google DeepMind, describes in an X post as an unbelievable amount of complex engineering to make AI feel instantaneous.Perhaps the most innovative technical addition is the introduction of thinking levels. Standardized across both the Flash-Lite and Pro variants, this feature allows developers to modulate the model’s reasoning intensity dynamically. For a simple classification task or a high-volume sentiment analysis, the model can be dialed down for maximum speed and minimum cost. Conversely, for complex code exploration, generating dashboards, or creating simulations, the thinking can be dialed up, allowing the model to perform deeper reasoning and logic before emitting its first response.Product: benchmarking the lite-weight heavy hitterWhile the “Lite” suffix often implies a significant sacrifice in capability, the performance data suggests a model that punches well into the territory of much larger systems. Gemini 3.1 Flash-Lite achieved an Elo score of 1432 on the Arena.ai Leaderboard, placing it in a competitive tier with models much larger in parameter count.Key benchmark results highlight its specialized strengths across diverse cognitive domains:Scientific knowledge: 86.9 percent on GPQA Diamond.Multimodal understanding: 76.8 percent on MMMU-Pro.Multilingual Q&A: 88.9 percent on MMMLU.Parametric knowledge: 43.3 percent on SimpleQA Verified.Abstract reasoning: 16.0 percent on Humanity’s Last Exam (full set)The model is particularly adept at structured output compliance—a critical requirement for enterprise developers who need AI to generate valid JSON, SQL, or UI code that won’t break downstream systems. In benchmarks like LiveCodeBench, Flash-Lite scored a 72.0 percent, outperforming several rivals in its weight class, including GPT-5 mini, which scored 80.4 percent on a different subset but lagged significantly in speed and cost efficiency. Furthermore, its performance on CharXiv Reasoning (73.2 percent) and Video-MMMU (84.8 percent) demonstrates that its multimodal capabilities are robust enough for complex chart synthesis and knowledge acquisition from video.The intelligence hierarchy: Flash-Lite vs. 3.1 ProTo understand Flash-Lite’s place in the market, one must look at it alongside Gemini 3.1 Pro, which Google released in mid-February 2026 to retake the AI crown. While Flash-Lite is the reflexes of the Gemini system, 3.1 Pro is undoubtedly the brain. The primary differentiator is the depth of cognitive processing. Gemini 3.1 Pro was engineered to double the reasoning performance of the previous generation, achieving a verified score of 77.1 percent on ARC-AGI-2—a benchmark designed to test a model’s ability to solve entirely new logic patterns it has not encountered during training.While Flash-Lite holds its own in scientific knowledge at 86.9 percent, the Pro model pushes that boundary to a staggering 94.3 percent, making it the superior choice for deep research and high-stakes synthesis. The application focus also differs significantly based on these reasoning gaps. Gemini 3.1 Pro is capable of vibe-coding—generating animated SVGs and complex 3D simulations directly from text prompts. For example, in one demonstration, Pro coded a complex 3D starling murmuration that users could manipulate via hand-tracking. It can even reason through abstract literary themes, such as translating the atmospheric tone of Emily Brontë’s Wuthering Heights into a functional web design.Gemini 3.1 Flash-Lite, conversely, is the workhorse for high-volume execution. It handles the millions of daily tasks—translation, tagging, and moderation—that require consistent, repeatable results without the massive compute overhead of a reasoning-heavy model. It fills a wireframe with hundreds of products instantly or orchestrates intent routing with 94 percent accuracy, as reported by early testers.1/8th the cost of the flagship Gemini 3.1 Pro model (and cheaper than its predecessor, Flash-Lite 2.5)For enterprise technical decision-makers, the most compelling part of the Gemini 3.1 series is the reasoning-to-dollar ratio. Google has priced Gemini 3.1 Flash-Lite at $0.25 per 1 million input tokens and $1.50 per 1 million output tokens.This pricing makes it significantly more affordable than competitors like Claude 4.5 Haiku, which is priced at $1.00 per 1 million input and $5.00 per 1 million output tokens. Even compared to Gemini 2.5 Flash, which cost $0.30 per 1 million input, Flash-Lite offers a cost reduction alongside its performance gains.When contrasted with Gemini 3.1 Pro—which maintains a price of $2.00 per million input tokens for prompts up to 200k—the strategic advantage of the dual-model approach becomes clear. In high-context usage (above 200,000 tokens per interaction), Flash-Lite is actually between 12x and 16x cheaper.ModelInputOutputTotal CostSourceQwen 3 Turbo$0.05$0.20$0.25Alibaba CloudQwen3.5-Flash$0.10$0.40$0.50Alibaba Clouddeepseek-chat (V3.2-Exp)$0.28$0.42$0.70DeepSeekdeepseek-reasoner (V3.2-Exp)$0.28$0.42$0.70DeepSeekGrok 4.1 Fast (reasoning)$0.20$0.50$0.70xAIGrok 4.1 Fast (non-reasoning)$0.20$0.50$0.70xAIMiniMax M2.5$0.15$1.20$1.35MiniMaxGemini 3.1 Flash-Lite$0.25$1.50$1.75GoogleMiniMax M2.5-Lightning$0.30$2.40$2.70MiniMaxGemini 3 Flash Preview$0.50$3.00$3.50GoogleKimi-k2.5$0.60$3.00$3.60MoonshotGLM-5$1.00$3.20$4.20Z.aiERNIE 5.0$0.85$3.40$4.25BaiduClaude Haiku 4.5$1.00$5.00$6.00AnthropicQwen3-Max (2026-01-23)$1.20$6.00$7.20Alibaba CloudGemini 3 Pro (≤200K)$2.00$12.00$14.00GoogleGPT-5.2$1.75$14.00$15.75OpenAIClaude Sonnet 4.5$3.00$15.00$18.00AnthropicGemini 3 Pro (>200K)$4.00$18.00$22.00GoogleClaude Opus 4.6$5.00$25.00$30.00AnthropicGPT-5.2 Pro$21.00$168.00$189.00OpenAIBy using a cascading architecture, an enterprise can use 3.1 Pro for the initial complex planning, architectural design, and deep logic, then hand off high-frequency, repetitive execution to Flash-Lite at one-eighth of the cost.This shift effectively moves AI from an expensive experimental cost center to a utility-grade resource that can be run over every log file, email, and customer chat without exhausting the cloud budget.Community and developer reactionsEarly feedback from Google’s partner network suggests that the 3.1 series is successfully filling a critical gap in the market for reliable autonomy. Andrew Carr, Chief Scientist at Cartwheel, has tested both models and noted their unique strengths. Regarding 3.1 Pro, he highlighted its substantially improved understanding of 3D transformations, which resolved long-standing rotation order bugs in animation pipelines. However, he found Flash-Lite to be a different kind of unlock for the business: “3.1 Flash-Lite is a remarkably competent model. It is lightning fast, but still somehow finds a way to follow all instructions… The intelligence to speed ratio is unparalleled in any other model”.For consumer-facing applications, the low latency of Flash-Lite has been the key to market expansion. Kolby Nottingham, Head of AI at Latitude, shared that the model achieved a 20 percent higher success rate and 60 percent faster inference times compared to their previous model, enabling sophisticated storytelling to a much wider audience than would have otherwise been possible. Reliability in data tagging has also emerged as a standout feature. Bianca Rangecroft, CEO of Whering, reported that by integrating 3.1 Flash-Lite into their classification pipeline, they achieved 100 percent consistency in item tagging, providing a highly reliable foundation for their label assignment and increasing confidence in structured outputs.Kaan Ortabas, Co-Founder of HubX, noted that as a root orchestration engine, Flash-Lite delivered sub-10 second completions with near-instant streaming and 97 percent structured output compliance. On the flagship side, Vladislav Tankov, Director of AI at JetBrains, noted a 15 percent quality improvement in the Pro model, emphasizing that it is stronger, faster, and more efficient, requiring fewer output tokens to achieve its goals.Licensing and enterprise availabilityBoth Gemini 3.1 Flash-Lite and Pro are offered through Google AI Studio and Vertex AI. As proprietary models, they follow a standard commercial software-as-a-service model rather than an open-source license. Operating through Vertex AI provides grounded reasoning within a secure perimeter, ensuring that high-volume workloads—like those being run by Databricks to achieve best-in-class results on the OfficeQA benchmark—remain protected by enterprise-grade security and data residency guarantees.However, they also are limited in terms of customizability and require persistent internet connectivity, as opposed to purely open source rivals like the powerful new Qwen3.5 series released by Alibaba over the last few weeks.The current preview status for Flash-Lite allows Google to refine safety and performance based on real-world developer feedback before general availability. For developers already building via the Gemini API, the transition to 3.1 Pro and Flash-Lite represents a direct performance upgrade at the same or lower price points, effectively lowering the barrier to entry for complex agentic workflows.The verdict: the new standard for utility AIThe release of Gemini 3.1 Flash-Lite represents the final piece of a strategic pivot for Google. While the industry has been obsessed with state-of-the-art reasoning for the most complex problems, the vast majority of enterprise work consists of high-volume, repetitive, but high-precision tasks. By providing both the brain in Gemini 3.1 Pro and the reflexes in Gemini 3.1 Flash-Lite, Google is signaling that the next phase of the AI race will be won by models that can think through a problem, but also execute that solution at scale.For the CTO or technical lead deciding which model to bake into their 2026 product roadmap, the Gemini 3.1 series offers a compelling argument: you no longer have to pay a reasoning tax to get reliable, instantaneous results. As Flash-Lite rolls out in preview today, the message to the developer community is clear: the barrier to intelligence at scale hasn’t just been lowered—it’s been dismantled.
OpenAI’s AI data agent, built by two engineers, now serves thousands of employees — and the company says anyone can replicate it
When an OpenAI finance analyst needed to compare revenue across geographies and customer cohorts last year, it took hours of work — hunting through 70,000 datasets, writing SQL queries, verifying table schemas. Today, the same analyst types a plain-English question into Slack and gets a finished chart in minutes.The tool behind that transformation was built by two engineers in three months. Seventy percent of its code was written by AI. And it is now used by thousands of OpenAI’s employees every day — making it one of the most aggressive deployments of an AI data agent inside any company, anywhere.In an exclusive interview with VentureBeat, Emma Tang, the head of data infrastructure at OpenAI whose team built the agent, offered a rare look inside the system — how it works, how it fails, and what it signals about the future of enterprise data. The conversation, paired with the company’s blog post announcing the tool, paints a picture of a company that turned its own AI on itself and discovered something that every enterprise will soon confront: the bottleneck to smarter organizations isn’t better models. It’s better data.”The agent is used for any kind of analysis,” Tang said. “Almost every team in the company uses it.”A plain-English interface to 600 petabytes of corporate dataTo understand why OpenAI built this system, consider the scale of the problem. The company’s data platform spans more than 600 petabytes across 70,000 datasets. Even locating the correct table can consume hours of a data scientist’s time. Tang’s Data Platform team — which sits under infrastructure and oversees big data systems, streaming, and the data tooling layer — serves a staggering internal user base. “There are 5,000 employees at OpenAI right now,” Tang said. “Over 4,000 use data tools that our team provides.”The agent, built on GPT-5.2 and accessible wherever employees already work — Slack, a web interface, IDEs, the Codex CLI, and OpenAI’s internal ChatGPT app — accepts plain-English questions and returns charts, dashboards, and long-form analytical reports. In follow-up responses shared with VentureBeat on background, the team estimated it saves two to four hours of work per query. But Tang emphasized that the larger win is harder to measure: the agent gives people access to analysis they simply couldn’t have done before, regardless of how much time they had.”Engineers, growth, product, as well as non-technical teams, who may not know all the ins and outs of the company data systems and table schemas” can now pull sophisticated insights on their own, her team noted.From revenue breakdowns to latency debugging, one agent does it allTang walked through several concrete use cases that illustrate the agent’s range. OpenAI’s finance team queries it for revenue comparisons across geographies and customer cohorts. “It can, just literally in plain text, send the agent a query, and it will be able to respond and give you charts and give you dashboards, all of these things,” she said.But the real power lies in strategic, multi-step analysis. Tang described a recent case where a user spotted discrepancies between two dashboards tracking Plus subscriber growth. “The data agent can give you a chart and show you, stack rank by stack rank, exactly what the differences are,” she said. “There turned out to be five different factors. For a human, that would take hours, if not days, but the agent can do it in a few minutes.”Product managers use it to understand feature adoption. Engineers use it to diagnose performance regressions — asking, for instance, whether a specific ChatGPT component really is slower than yesterday, and if so, which latency components explain the change. The agent can break it all down and compare prior periods from a single prompt.What makes this especially unusual is that the agent operates across organizational boundaries. Most enterprise AI agents today are siloed within departments — a finance bot here, an HR bot there. OpenAI’s cuts horizontally across the company. Tang said they launched department by department, curating specific memory and context for each group, but “at some point it’s all in the same database.” A senior leader can combine sales data with engineering metrics and product analytics in a single query. “That’s a really unique feature of ours,” Tang said.How Codex solved the hardest problem in enterprise dataFinding the right table among 70,000 datasets is, by Tang’s own admission, the single hardest technical challenge her team faces. “That’s the biggest problem with this agent,” she said. And it’s where Codex — OpenAI’s AI coding agent — plays its most inventive role.Codex serves triple duty in the system. Users access the data agent through Codex via MCP. The team used Codex to generate more than 70% of the agent’s own code, enabling two engineers to ship in three months. But the third role is the most technically fascinating: a daily asynchronous process where Codex examines important data tables, analyzes the underlying pipeline code, and determines each table’s upstream and downstream dependencies, ownership, granularity, join keys, and similar tables.”We give it a prompt, have Codex look at the code and respond with what we need, and then persist that to the database,” Tang explained. When a user later asks about revenue, the agent searches a vector database to find which tables Codex has already mapped to that concept.This “Codex Enrichment” is one of six context layers the agent uses. The layers range from basic schema metadata and curated expert descriptions to institutional knowledge pulled from Slack, Google Docs, and Notion, plus a learning memory that stores corrections from previous conversations. When no prior information exists, the agent falls back to live queries against the data warehouse.The team also tiers historical query patterns. “All query history is everybody’s ‘select star, limit 10.’ It’s not really helpful,” Tang said. Canonical dashboards and executive reports — where analysts invested significant effort determining the correct representation — get flagged as “source of truth.” Everything else gets deprioritized.The prompt that forces the AI to slow down and thinkEven with six context layers, Tang was remarkably candid about the agent’s biggest behavioral flaw: overconfidence. It’s a problem anyone who has worked with large language models will recognize.”It’s a really big problem, because what the model often does is feel overconfident,” Tang said. “It’ll say, ‘This is the right table,’ and just go forth and start doing analysis. That’s actually the wrong approach.”The fix came through prompt engineering that forces the agent to linger in a discovery phase. “We found that the more time it spends gathering possible scenarios and comparing which table to use — just spending more time in the discovery phase — the better the results,” she said. The prompt reads almost like coaching a junior analyst: “Before you run ahead with this, I really want you to do more validation on whether this is the right table. So please check more sources before you go and create actual data.”The team also learned, through rigorous evaluation, that less context can produce better results. “It’s very easy to dump everything in and just expect it to do better,” Tang said. “From our evals, we actually found the opposite. The fewer things you give it, and the more curated and accurate the context is, the better the results.”To build trust, the agent streams its intermediate reasoning to users in real time, exposes which tables it selected and why, and links directly to underlying query results. Users can interrupt the agent mid-analysis to redirect it. The system also checkpoints its progress, enabling it to resume after failures. And at the end of every task, the model evaluates its own performance. “We ask the model, ‘how did you think that went? Was that good or bad?'” Tang said. “And it’s actually fairly good at evaluating how well it’s doing.”Guardrails that are deliberately simple — and surprisingly effectiveWhen it comes to safety, Tang took a pragmatic approach that may surprise enterprises expecting sophisticated AI alignment techniques.”I think you just have to have even more dumb guardrails,” she said. “We have really strong access control. It’s always using your personal token, so whatever you have access to is only what you have access to.”The agent operates purely as an interface layer, inheriting the same permissions that govern OpenAI’s data. It never appears in public channels — only in private channels or a user’s own interface. Write access is restricted to a temporary test schema that gets wiped periodically and can’t be shared. “We don’t let it randomly write to systems either,” Tang said.User feedback closes the loop. Employees flag incorrect results directly, and the team investigates. The model’s self-evaluation adds another check. Longer term, Tang said, the plan is to move toward a multi-agent architecture where specialized agents monitor and assist each other. “We’re moving towards that eventually,” she said, “but right now, even as it is, we’ve gotten pretty far.”Why OpenAI won’t sell this tool — but wants you to build your ownDespite the obvious commercial potential, OpenAI told VentureBeat that the company has no plans to productize its internal data agent. The strategy is to provide building blocks and let enterprises construct their own. And Tang made clear that everything her team used to build the system is already available externally.”We use all the same APIs that are available externally,” she said. “The Responses API, the Evals API. We don’t have a fine-tuned model. We just use 5.2. So you can definitely build this.”That message aligns with OpenAI’s broader enterprise push. The company launched OpenAI Frontier in early February, an end-to-end platform for enterprises to build and manage AI agents. It has since enlisted McKinsey, Boston Consulting Group, Accenture, and Capgemini to help sell and implement the platform. AWS and OpenAI are jointly developing a Stateful Runtime Environment for Amazon Bedrock that mirrors some of the persistent context capabilities OpenAI built into its data agent. And Apple recently integrated Codex directly into Xcode.According to information shared with VentureBeat by OpenAI, Codex is now used by 95% of engineers at OpenAI and reviews all pull requests before they’re merged. Its global weekly active user base has tripled since the start of the year, surpassing one million. Overall usage has grown more than fivefold.Tang described a shift in how employees use Codex that transcends coding entirely. “Codex isn’t even a coding tool anymore. It’s much more than that,” she said. “I see non-technical teams use it to organize thoughts and create slides and to create daily summaries.” One of her engineering managers has Codex review her notes each morning, identify the most important tasks, pull in Slack messages and DMs, and draft responses. “It’s really operating on her behalf in a lot of ways,” Tang said.The unsexy prerequisite that will determine who wins the AI agent raceWhen asked what other enterprises should take away from OpenAI’s experience, Tang didn’t point to model capabilities or clever prompt engineering. She pointed to something far more mundane.”This is not sexy, but data governance is really important for data agents to work well,” she said. “Your data needs to be clean enough and annotated enough, and there needs to be a source of truth somewhere for the agent to crawl.”The underlying infrastructure — storage, compute, orchestration, and business intelligence layers — hasn’t been replaced by the agent. It still needs all of those tools to do its job. But it serves as a fundamentally new entry point for data intelligence, one that is more autonomous and accessible than anything that came before it.Tang closed the interview with a warning for companies that hesitate. “Companies that adopt this are going to see the benefits very rapidly,” she said. “And companies that don’t are going to fall behind. It’s going to pull apart. The companies who use it are going to advance very, very quickly.”Asked whether that acceleration worried her own colleagues — especially after a wave of recent layoffs at companies like Block — Tang paused. “How much we’re able to do as a company has accelerated,” she said, “but it still doesn’t match our ambitions, not even one bit.”
Wall Street’s ‘fear gauge’ is rising as Iran conflict escalates. Here’s what investors should watch.
Wall Street’s “fear gauge” was rising on Tuesday as the S&P 500 and other major U.S. equity indexes touched their lowest levels of the year, before an afternoon rebound caused them to recoup much of their earlier losses.