BUSINESS

“KPop Demon Hunters” gets a new role-playing item with Rumi’s Sain-Geom — aka the “Four Tiger Sword” — with premium electronic and interactive features, Hasbro announced at San Diego Comic-Con on Thursday.

Trump had 34% and 39% approval ratings in the latest two polls released Thursday.

Trump said Iran hasn’t “received enough pain yet.”

The United Kingdom is seeking to extradite Andrew Tate and his brother Tristan, whom British authorities have charged with dozens of alleged crimes.

UnitedHealth Group (UNH) just delivered the quarter battered shareholders have been waiting for.The health care company blew past Wall Street’s profit prediction, increased its most widely monitored cost index, and boosted its full-year earnings outlook. The results seemed to vindicate investors who had bought the stock during the collapse of confidence in the company.But UnitedHealth’s recovery isn’t over.The corporation continues to face federal review of its Medicare billing methods, and management has warned that pressure on medical costs will remain. This leaves investors with a tough question: Is UnitedHealth reverting to its old earnings machine, or is the stock rallying because the biggest risk is now out of the way?Investor Steve Weiss has continued to buy UnitedHealth, calling it a “permanent compounder” on CNBC. Billionaire investor Bill Ackman had the opposite view, warning the company’s mounting problems could expose deeper trouble.Weiss has recent earnings on his side, while the government probes prohibit investors from giving up on Ackman.UnitedHealth’s earnings destroyed the bearish operating caseUnitedHealth reported second-quarter adjusted earnings of $6.38 a share, according to Reuters, well above the $4.90 analysts were expecting. Revenue was $112 billion, above consensus and up a little over last year.The company also boosted its projection for 2026 adjusted earnings to a range of $19.50 to $20 per share, from a previous view of at least $18.25.Related: JPMorgan resets UnitedHealth stock target for 2026Earnings were not the most important number; it was the medical loss ratio for UnitedHealth.The ratio, which tracks the percentage of premium revenue used for members’ medical treatment, improved to 86.7% from 89.4% a year earlier and was well below expectations of analysts, Reuters noted. If health costs are lower, an insurer can retain a larger share of the premium revenue for administrative expenditures and profit.The improvement shows that UnitedHealth’s plan redesigns, pricing measures, and medical-management strategies are starting to take hold.Optum offered another positive indicator. Operating income in the health services unit soared 29% to almost $4 billion, Reuters confirmed, quashing fears that the problems had infected more than the insurance business.

UnitedHealth may be recovering faster than investors can price the risk.Nagle/Bloomberg via Getty Images

One number makes UnitedHealth’s recovery less reassuringUnitedHealth’s 86.7% medical care ratio comprised about $860 million of favorable prior-period development. That indicates some of the improvement came from the corporation modifying projections for medical claims from earlier periods, not just from decreased costs incurred during the quarter.This does not make the result invalid. Still, investors should not expect all of the margin improvement to be a permanent change in the underlying business.Management also said medical cost trends remain elevated in Medicare and commercial insurance, Investors Business Daily noted, mainly driven by specialty pharmaceuticals and costs related to the No Surprises Act. UnitedHealth’s raised earnings outlook was still below some analyst projections, helping the stock give back much of its initial post-results rally.UnitedHealth is likewise cutting back on its less profitable Medicare Advantage plans. Such moves can restore margins, but there’s a trade-off: fewer members, and maybe slower revenue growth.So the essence of the turnaround is that UnitedHealth is making more money from a smaller, more worth-its-price membership.The threat UnitedHealth’s earnings cannot resolveThe Justice Department has investigated UnitedHealth’s Medicare Advantage billing practices, including how diagnoses sent to the government could affect payments.UnitedHealth said in 2025 that it was responding to official civil and criminal requests and said it had confidence in the integrity of its processes. An investigation doesn’t imply any wrongdoing, and you shouldn’t make any assumptions until the process is complete.But it cuts to the core of UnitedHealth’s business model.Insurers who offer Medicare Advantage plans receive paid extra for people with serious health issues. Governments immediately penalized for illegal payments resulting from coding practices could face financial penalties, operational modifications, or reputational damage.That’s the heart of the bearish thesis. It is not that UnitedHealth cannot improve one quarter’s medical costs. It is that the market still cannot confidently estimate the potential cost of federal scrutiny.Key takeaways for UnitedHealth investorsAdjusted earnings substantially exceeded Wall Street’s forecast.The medical care ratio improved to 86.7%.Optum’s operating income increased 29%.UnitedHealth raised its 2026 earnings outlook.Some cost improvement came from favorable prior-period development.Federal Medicare scrutiny remains unresolved.UnitedHealth already answered the first question investors asked after its crash — namely, whether the corporation can still make a lot of money.It has not answered the more critical question: whether the practices, laws, and medical-cost assumptions enabling those earnings will stay in place. That makes UnitedHealth’s recovery especially difficult to value.The operating recuperation is real. The momentum in earnings is picking up. And the latest revelation weakened the argument that the business is permanently broken.But one good quarter doesn’t close a federal inquiry.UnitedHealth’s stock could continue to climb as margins improve. The danger is that investors are pricing the turnaround faster than anyone can price what investigators may eventually find.Related: UnitedHealth CFO sends stark warning after earnings

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 dealBackyard barbecues during summer are an American tradition. Inviting friends and family, throwing some delicious food on the cooktop, and having a great time with loved ones is something that many people look forward to every season. The more convenient it is to host a barbecue, the more likely it is to happen, and if you’re someone who loves having friends over or just cooking outdoors for yourself or the family, having the right tools can streamline the whole process, allowing you to spend more time having fun and less time worrying and cleaning. If you’ve been shopping for some grill accessories, chances are you’ve seen Blackstone pretty much everywhere. They’re versatile, easy to clean, and can make your outdoor feast even more delicious, and they also make great accessories. The Blackstone Deluxe Stainless Steel Griddle Spatula Set is a great addition to any outdoor cooking setup, and the best part is, this set is just $24 at Walmart. Blackstone Deluxe Stainless Steel Griddle Spatula Set, $24 (was $30) at Walmart

Courtesy of Walmart

Shop at WalmartWhy do shoppers love it?This set includes an extra-large hamburger spatula that’s great for smashing, flipping, and serving, two griddle spatulas with a beveled edge, stainless steel tongs with a locking mechanism, a paintbrush-style griddle scraper to make cleaning easier, and a 32-ounce squeeze bottle with a twist-top nozzle for flavoring, steaming, and cleaning. It also features a “Griddle Essentials” cookbook for inspiration and ideas. Each tool is designed with heavy-gauge brushed stainless steel and plastic handles for durability, offering professional-grade accessories that are perfect for cooking burgers, pancakes, vegetables, eggs, and tons of other foods. They’re useful for Blackstone griddle cooktops, but can also be used to flip burgers on a normal grill, and even stainless steel kitchen pans. Related: Walmart has a Blackstone griddle that’s ‘built like a tank’ for under $100, and it’s selling fastThis set allows you to cook seamlessly, keeping food moving across the cooking surface, adding water or oil when needed, and transitioning from breakfast to dinner without missing a beat. If you’ve ever cooked breakfast for a crowd or tried to keep burgers, onions, and toasted buns going at the same time, you know that having the right tools to help everything go smoothly is a huge advantage. Then, with the tools already on hand, it’s just as easy to clean up with the griddle scraper and squeeze bottle at the end of the night. Details to knowItems: This set includes an extra-large hamburger spatula, two griddle spatulas, and a pair of tongs. It also has a griddle scraper and a 32-ounce squeeze bottle. Material: The stainless steel material works well and can last for decades. Bonus cookbook: Enjoy a bonus cookbook with this set to help kick off the backyard season. One reviewer loved this set, saying “These are made of heavy stainless steel. I’ve had them over high heat plenty of times and haven’t seen any warping. The set comes with a wide burger spatula that has a sharp, beveled edge. If you like making smashburgers, this is key for getting under that crust without tearing the meat apart. The handles also have built-in thumb rests and hooks on the end. It sounds small, but when you’re juggling three different things in the heat, having a spot to hang your greasy tools is a lifesaver.””They are very useful, sturdy, and beautiful as well,” said another shopper. “I love them so much I purchased a set for each of my adult children as nice gifts.”Shop more dealsBlackstone 25-Piece Ultimate Griddle Set, $60 (was $69) at WalmartBlackstone Signature 12-Inch Round Basting Cover, $20 (was $26) at WalmartMibote Griddle Accessory Kit, $32 (was $53) at WalmartThe Blackstone Deluxe Stainless Steel Griddle Spatula Set offers an affordable, accessible option for any cooktop grilling this season. Whether it’s Saturday morning breakfast or weeknight dinner for the family, this set has you covered for just $24.

The post, now deleted, touted a token called $VLAD and falsely claimed it would be listed on Robinhood.

On July 21, Mizuho told clients that Oracle (ORCL) stock could more than double from here.The bank reiterated its Outperform rating and kept its $320 price target, a call that implies about a 164% rise from Monday’s close.That number stands out because Oracle has been one of 2026’s most punished large-cap tech names.ORCL shares are down about 38% from the start of the year and more than 48% over the past 12 months.So when a major bank says a beaten-down stock can double, the first question worth asking is simple. What does Mizuho see that the bears do not?Why Mizuho is calling Oracle stock a discount buyMizuho analyst Siti Panigrahi framed the case around one word: valuation.Oracle now trades at about 14 times its expected fiscal 2027 adjusted earnings, a level Panigrahi called a clear discount compared to its peers, TipRanks reported.He believes Oracle offers one of the most attractive risk-reward setups.That argument rests on a gap. Oracle is growing faster than many rivals, yet the stock trades cheaper than them.Panigrahi pointed to steadier execution in three areas: converting signed contracts into live capacity, the quality of the backlog, and clearer visibility into how Oracle funds its buildout.Mizuho is arguing that the market has priced Oracle as a troubled spender while ignoring the revenue already under contract.

Oracle has become one of the AI data-center race’s biggest spenders, and its stock has paid the price in 2026.NurPhoto / Getty Images

The backlog Oracle is counting on to grow intoOracle finished fiscal 2026 with remaining performance obligations of $638 billion, up 363% compared to last year.Remaining performance obligations, or RPO, is simply the value of signed contracts a company has not yet delivered.In plain terms, it is revenue Oracle has already booked but still has to build the capacity to serve.That backlog is why Oracle’s management predicts about 34% total revenue growth in fiscal 2027, up from 17% in fiscal 2026.More AI Infrastructure Stocks:UBS says to ‘buy the dip’ in Bloom Energy stockNvidia CEO doubles down on AI and stock market verdictCathie Wood buys $2.1M of tumbling AI stockOracle also signed $67 billion in new AI infrastructure contracts in the fourth quarter alone, much of it prepaid or built on customer-owned hardware.Prepaid deals matter because customers front the cash, which eases the pressure on Oracle’s own wallet.Why Oracle stock fell so far in the first placeThe fear driving the selloff is cost. Building AI data centers is very expensive, and Oracle is spending like few companies ever have.Oracle estimated roughly $90 billion to $95 billion in gross capital spending for fiscal 2027, MLQ reported.Free cash flow ran deeply negative in fiscal 2026 as that buildout accelerated.Here is what investors have been cautious of:A credit profile pushed to BBB-, one grade above junkA capital need that CLSA estimates could reach $500 billion through 2030Roughly $40 billion in planned debt and equity raises for fiscal 2027The picture worsened this week when a New Mexico regulator again rejected the pipeline meant to power Oracle’s flagship Project Jupiter data center, threatening its timeline.How Mizuho’s target stacks up against the rest of Wall StreetA 164% target sounds ambitious until you see where the rest of Wall Street sits.Mizuho is the most bullish voice, but it is not alone in seeing a large rise from these levels.Of the 32 analysts covering Oracle, 28 rate it a Buy or Strong Buy, while four gave a Hold rating. The 12-month average target sits at $259.76.Related: Oracle stock makes rattling move after major setbackThe difference between the average target and today’s price is a signal that Wall Street mostly believes Oracle is worth far more than it trades for. Most analysts want to see a debt or equity raise land cleanly before pricing in the full recovery. Mizuho is willing to price it in now, ahead of that confirmation.How ORCL has moved against the marketOracle has badly trailed the wider market during this stretch, and that shows how much sentiment has turned.Here is the recent scorecard for ORCL:Past five days: down about 4%, near $125.73Past month: down about 28%Year to date: down about 38%52-week range: a low of $120.03 against a high near $346That 52-week range is what Mizuho is leaning on. A stock trading close to $120 with a backlog of over $600 billion is either a value trap or a rare mispricing.What Oracle investors should watch before actingA double-your-money target is exciting, but it depends on several things going right.Before treating Mizuho’s call as a green light, track the catalysts that actually decide the final outcome:Financing clarity. Any large debt or equity raise will show how Oracle plans to close its funding gap, and could pressure shares near term.Power and permits. A resolution on Project Jupiter’s pipeline and air quality permit would remove a concrete delay risk.The next earnings check. Oracle reports first-quarter fiscal 2027 results on September 10, the next real test of whether the backlog is converting.Credit re-ratings. Mizuho flags potential upgrades as a mid-term catalyst, so watch the rating agencies.The bull case is real. Oracle has the demand, the backlog, and a cheaper multiple than its peers.The risk is equally real. A company funding a historic buildout largely on borrowed money has little room for error, and the market is pricing that in.Nothing here is a recommendation to buy or sell. Weigh the growth against the balance sheet and decide what fits your own risk tolerance.Related: Analyst sends chilling Oracle stock verdict

Black Forest Labs (BFL) is expanding its FLUX family beyond image generation with today’s launch of FLUX 3, a multimodal frontier model trained to understand and generate images, or combined audio/video clips up to 20 seconds from a single prompt — and to extend the same underlying architecture to robotic vision and actions.The Freiburg, Germany-based AI lab says FLUX 3 is jointly trained across those modalities rather than assembling separate image, video and audio models behind a common interface. That distinction is central to the company’s pitch: BFL wants enterprises to think about creative generation, simulation, computer use and robotics as connected applications of a single capability it calls visual intelligence — models, in the company’s words, “that can perceive, predict, and act across physical and digital environments.” This release marks BFL’s first public video generation model. FLUX 3 will be offered through four product lines: FLUX 3 Video, FLUX 3 Image, FLUX 3 Action and the upcoming, open source FLUX 3 Dev. FLUX 3 Video, with optional native audio generation, and FLUX 3 Action are entering a gated “Early Access” program now, to which anyone can apply, but which BFL must approve. There is presently no public access through BFL’s application programming interface (API) or those of partners yet, but the company says FLUX 3 Image will roll out in the coming weeks, followed by general availability. The limited initial availability rollout echoes the release strategies of new models from other frontier labs in the U.S. lately, including Anthropic and OpenAI, though those were ostensibly for security concerns and due to government request. What the company has not announced is pricing, production service-level commitments, evaluation methodology, sample sizes, rater counts or any image-model benchmarks at all. Enterprise buyers therefore cannot yet calculate total cost of ownership or independently reproduce the video comparisons.Another big notable omission: FLUX 3 is not launching with downloadable weights at this time, nor an open source license. BFL says faster and open-weight versions will arrive later this year, and its technical blog names FLUX 3 Dev as “open-weight access to a multimodal backbone, for content creation (video, audio and image) and action prediction” — a considerably broader commitment than any previous FLUX Dev release, all of which covered images only.But it arrives last in the sequence. Developers accustomed to receiving a locally deployable FLUX variant alongside — or soon after — a major model announcement will have to wait. That delay does not negate the company’s commitment, but it is disappointing given the role open weights have played in FLUX’s adoption thus far. Flux 3 is rated higher than the competition, but missing pricing and benchmarking details may prevent rapid enterprise adoptionBFL has published several benchmark comparisons, but they’re qualified as preliminary — with full benchmark results and methodology to be published later during broader general availability. In early head-to-head preference testing on 10-second, 720p text-to-video clips with audio, the company says FLUX 3 was preferred over Luma Ray 3.2 in 93% of comparisons, Runway Gen-4.5 in 77%, Grok Imagine Video in 69%, Kling v3 Pro in 60%, Happy Horse v1 in 59%, Happy Horse 1.1 in 57%, and both Seedance 2.0 and Google’s Gemini Omni Flash in 52%.One caveat travels with every one of those figures, and it comes from BFL itself. The chart carrying the results is labeled a “preliminary evaluation of an early FLUX 3 candidate” — meaning the numbers describe a pre-release checkpoint rather than the model now entering early access. That cuts both ways: the shipping model may perform better, but nothing published today measures what customers will actually call.Luma Ray 3.2 and Runway Gen-4.5, where FLUX 3 posted 93% and 77%, are the softest comparisons on the list — established products, but not the models currently setting the pace in independent video rankings. Those are real wins, and they are the ones least likely to change an enterprise shortlist.Seedance 2.0, at 52%, is a statistical coin flip against a model most Western enterprises cannot currently procure. ByteDance indefinitely postponed Seedance 2.0’s international rollout after Netflix, Warner Bros., Disney, Paramount and Sony sent legal threats over alleged systematic copyright infringement, and that suspension remains in place. Tying a frozen product is neither a strong claim nor a damaging one.Gemini Omni Flash, also at 52%, matters much more. Omni is the closest large-platform analogue to what FLUX 3 is attempting — multimodal input, video and audio-aware creation, conversational editing — and by BFL’s own measurement, the two are indistinguishable on 10-second text-to-video quality. Google’s advantage in that matchup is that Omni is generally available via Google’s Gemini API for $0.10 per second of generated 720p video, or a 10-second clip for around.One regional wrinkle matters for a German company’s home market. Editing uploaded video is unavailable to Omni Flash users in the European Economic Area, Switzerland and the United Kingdom, though editing video the model itself generated is permitted. A European enterprise that wants to run its existing footage through a generative editing pass cannot currently do so on Omni Flash.Here’s a rough guide for enterprises considering which video models to rely upon: ModelMax single-generation durationMax resolutionKey constraintsPrice per 10-second clip (720p)Price per 10-second clip (1080p)Price per 10-second clip (4K)FLUX 3 Video 20 seconds Not stated; evaluations run at 720p Early access; no published SLA or pricing Not announced Not announced Not announced HappyHorse 1.1 15 seconds 1080p No 4K; closed weights Not published (v1.0 reseller rate is ~$1.82) Not published (v1.0 reseller rate is ~$3.12) n/a Veo 3.1 Per-second billing 4K Supports clip extension; preview $4.00 $4.00 $6.00 Veo 3.1 Fast Per-second billing 4K Preview $1.00 $1.20 $3.00 Veo 3.1 Lite Per-second billing 1080p No 4K, no clip extension; preview $0.50 $0.80 n/a Gemini Omni Flash 10 seconds (3s minimum) 720p at 24 FPS Preview abd no EU access$1.00 n/a n/a One architecture for media generation and physical actionFLUX 3 builds on Self-Flow, BFL’s method for aligning multimodal understanding and generation within one architecture, publicized back in March 2026. The company says it significantly scaled up compute and data to train across video, images and audio simultaneously, and that testing showed video generation and action prediction do not require separate foundations — the same architecture could be extended to action prediction without sacrificing what it learned from video.”We place vision at the center of our approach because it is the most signal-rich medium of the physical world. Images convey structure, images and video teach spatial relationships, video teaches dynamics, and actions reveal causal relationships. But vision alone is not the complete picture,” said Robin Rombach, co-founder and CEO of BFL, in a pre-release statement provided to VentureBeat. “True intelligence means perceiving the world: predicting how it will change, taking action, and learning from the results. Joint training within one unified architecture is what will get us there, because each training modality strengthens the others. Audio conveys timing, prosody, and physical events that elude vision. Language conveys goals, abstractions, and instructions that pixels cannot easily express.”He put the case more bluntly elsewhere in the announcement: “You can’t cheat reality. A model that only learns images can only generate images. But the world is not made of still frames. It moves, sounds, changes, and responds.”BFL says FLUX 3 targets creative tooling, media, design, e-commerce and physical AI, supporting video generation with synchronized audio, precise image editing, product and material consistency across motion, multilingual generation and robotic action prediction. It is already being tested by Canva, Burda, Magnific (formerly Freepik), Krea and Picsart.For creative software companies, the appeal is consolidation. A single foundation could potentially support storyboarding, image editing, product rendering, video variation and localization without repeatedly translating assets and instructions between disconnected models.For robotics teams, the potential value is data efficiency. Models that already encode motion, object behavior and physical change may need less task-specific robot training than systems starting from raw demonstrations.What FLUX 3 Video can actually doThe video tier is the most concretely specified part of the launch, and it settles a question that had been circulating as rumor: FLUX 3 generates clips of up to 20 seconds with audio in a single generation. Every video output comes with native audio. For comparison, HappyHorse 1.0 tops out at 15 seconds of 1080p with synchronized audio — though BFL has not stated what resolution its 20-second clips run at, and its published evaluations were conducted at 720p. Still, a 20-second long clip from a single prompt is among the longest yet achieved, matching OpenAI’s discontinued Sora model.The capability list BFL published covers:Text-to-video generation.Image-to-video generation, either animating from a starting frame or using images as visual references.Video-to-video generation from a reference clip, carrying elements such as a specific character into a new scene or context.Generative video-audio continuation from existing video and audio input.Keyframe-to-video generation for controlled transitions between defined moments. Multilingual dialogue.A broad range of visual styles and aspect ratios, from candid camcorder footage to animation and cinematics.Typography generation and animated design.Agentic chaining of individual clips into longer, multi-shot sequences.That last item is the one enterprise video teams should look at hardest. BFL claims the capabilities combine to produce sequences lasting several minutes, with visual references keeping characters consistent across scenes. If that holds up under production conditions, it addresses the constraint that has kept generative video out of most commercial pipelines: not clip quality, but continuity across shots.It is also the capability where competition is most direct. HappyHorse 1.1’s headline upgrade is R2V, or Reference-to-Video, which accepts multiple character reference images to hold identity stable across generated footage — the same problem, approached at the input layer rather than through agentic clip chaining. Alibaba also claims zero-drift lip sync and has specifically targeted the artifacts that mark commercial AI video as synthetic, including facial oiliness and over-sharpening. Character consistency is where this category is being contested, and both companies know it.BFL says FLUX 3 Video is already particularly strong at human facial expressions, associating sounds with physical events, and multilingual output. On the image side, the company says preliminary evaluations conducted during midtraining show significant improvement over earlier FLUX versions in complex prompt handling and text generation, including high-accuracy text in multiple languages. It published no image benchmarks or win rates.FLUX-mimic tests whether video models can become robot modelsBFL is applying its unified-architecture thesis through FLUX-mimic, a video-action model built on FLUX 3 and developed with Swiss firm Mimic Robotics, one of the first partners to receive early access.The technical blog describes two distinct routes to action prediction: integrating native action prediction directly into FLUX 3, scaling up the initial Self-Flow work; and using the pretrained video backbone as a dynamics-aware foundation from which specialized action models can be finetuned with limited task-specific data. FLUX-mimic is the second route — the FLUX 3 backbone combined with mimic’s robot-learning and production-deployment expertise in dexterous manipulation.FLUX-mimic is designed for general-purpose robotic manipulation: helping robots understand a visual scene, predict the consequences of an action, and adapt to new tasks with far less task-specific data. BFL and Mimic Robotics say that depending on task difficulty, the model can be finetuned for a specific manipulation task with as little as 30 minutes of robot data, where prior approaches have required 30 or more hours.”The hardest part of robotics is data,” said Elvis Nava, CTO of Mimic Robotics, in a statement provided to VentureBeat. “Every new task normally means hours of a robot repeating itself. Because FLUX-mimic is built on top of frontier video models that already understand how the physical world behaves, it picks up a new task in minutes, not days. This way, we can leapfrog the current state of the art in robot learning.”BFL argues that a model trained only on images cannot understand a world that “moves, sounds, changes, and responds,” and that physical understanding is what produces convincing generated footage. Google makes a nearly identical claim for Gemini Omni. Its developer documentation cites “world knowledge” that combines “an understanding of physics” with Gemini’s grasp of history, science and cultural context. Its marketing is blunter still: “Most AI models just predict the next pixel to build a narrative or an image. Gemini Omni is different,” the company posted in June, crediting the model with “an intuitive understanding of forces like gravity, kinetic energy, and fluid dynamics for more realistic movements that follow real-world logic.” The practical consequence for enterprise buyers is that world-model language is not a differentiator. Two of the three leading video systems now market physical understanding as their central advantage, and neither has published a benchmark that measures it. There is no standard test for whether generated water behaves like water, whether a dropped object falls at a plausible rate, or whether a sound arrives when the impact does. Human preference ratings capture some of it indirectly. Nothing else on offer captures it at all.Open weights helped make FLUX an industry standardBFL officially launched in summer 2024 and gained a name for itself in the AI industry in the intervening two years for its commitment to open sourcing high-quality AI image models beloved by developers, creatives, and enterprises. The company’s founders, including Rombach, Andreas Blattmann and Patrick Esser, previously helped create VQGAN, latent diffusion and Stable Diffusion, the latter the open source technology that kicked off broad AI generation capabilities for the masses and currently used by many AI image generators and companies. That reach translated into commercial distribution. FLUX models now power generative features inside Adobe Photoshop, Picsart and Nous Research’s Hermes Agent, among other platforms, and the company cites film director Martin Scorsese among professional users.Wired magazine described Black Forest Labs as a relatively small company that nevertheless became a leading competitor to Silicon Valley’s largest AI labs, with FLUX models ranking near the top of image benchmarks and becoming some of the most downloaded text-to-image models on AI code sharing community Hugging Face. The company says it now runs a 100-person team across Freiburg and San Francisco.FLUX.1 Dev, FLUX.1 Kontext Dev, FLUX.1 Fill Dev and related control models, released shortly after the firm’s launch, gave researchers and creative-tool developers access to downloadable checkpoints, local inference and integrations with frameworks including Hugging Face Diffusers and ComfyUI. FLUX.1 Kontext Dev, for example, was released as an open-weight model for research and noncommercial use, with generated outputs permitted for commercial purposes under the applicable license.The company continued that pattern with FLUX.2 Dev in late 2025, a 32-billion-parameter open-weight model combining generation and multi-reference editing. Black Forest Labs called it the strongest open-weight image generation and editing model available at launch and released weights, reference inference code and optimized implementations for consumer Nvidia GPUs.FLUX 3 Dev raises the stakes on that evaluation. Previous Dev releases were image models. This one is described as a multimodal backbone spanning video, audio, image and action prediction — meaning a single license will govern whether a company can locally deploy a model that touches both content production and physical machinery. BFL hasn’t yet shared information about its license, the parameter count, quantizations or hardware requirements.The company frames open weights as an enterprise feature rather than a community gesture, arguing they enable secure, low-latency local deployment for applications like robotic control systems and let teams adapt FLUX 3 to their own data, products and workflows. The financial backing behind FLUX 3 is worth noting alongside the technical claims. Black Forest Labs is valued at $3.25 billion and has raised more than $450 million from investors including a16z, AMP, Salesforce Ventures, Nvidia, General Catalyst, Adobe Ventures, Figma Ventures, Canva and Deutsche Telekom’s T.Capital.

The former Signet developer aims to launch early next year a marketplace connecting stablecoin issuers with regional lenders to manage reserves.