BUSINESS

But for first-round upsets, the big AI platforms have some hot tips

As indictments mount, fans and stock investors are losing trust that only an integrity commissioner can restore.

Thursday afternoon’s matchup against Louisville will be the first time Larry Hodgson, who has dementia, will see his son in person leading a team as a head coach.

“Building reliable infrastructure across blockchain networks and traditional financial rails is hard — there are no shortcuts,” said Polymarket CEO Shayne Coplan.

Multiple women have recalled sexual abuse by Cesar Chavez, a prominant figure in the Latino civil rights movement.

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 dealHallway closets and kitchen cabinets offer convenient storage, but they can quickly become cluttered without a way to organize them. Messy spaces don’t just look unsightly — they bring unnecessary chaos to your day. A disorganized home doesn’t just make it harder to find the things you need, but it can also mean wasted money when snacks expire in the pantry, or you buy an extra of the product you previously overlooked. One simple solution for streamlining these spaces is with stackable storage drawers, and right now, Amazon’s no. 1 bestseller, the Vtopmart 5-Piece Stackable Storage Drawers Set, is 10% off. These transparent drawers are sturdy and versatile, and with this limited-time deal, instead of the regular price of $49, you can score them for just $44. These savings come ahead of Amazon’s Big Spring Sale and just in time for spring cleaning season. The set can help you spruce up your kitchen countertops, bathroom closets, laundry room cabinets, under-the-sink spaces, and more.Vtopmart 5-Piece Stackable Storage Drawers Set, $44 (was $49) at Amazon

Courtesy of Amazon

Shop at AmazonHow to use stackable storage drawersThere is endless potential for organizing your home with these stackable storage drawers. My own home is filled with a set of very similar plastic storage drawers, and I’ve utilized them in several spots. One drawer was placed on the shelf of my kitchen pantry as a go-to spot for my favorite snacks. In the bathroom, several drawers were used for cosmetics, nail polish, hair accessories, and skincare products. My hallway junk closet had a few drawers to organize my smaller knick-knacks. Even the entryway closet was given a drawer, where it held winter accessories, rain ponchos, and small umbrellas. The transparent design makes it ultra-convenient to find whatever you need, and the compact footprint can fit in numerous areas. The clean and modern design is attractive enough to be out in the open, like on the bathroom counter, but the drawers are small enough to fit in tight spaces, like the office bookshelf.Details to know Pieces in the set: The five-piece set includes two small drawers, two medium drawers, and one wide drawer.Material: Shatter-resistant plastic.Are they dishwasher safe?: No. If these drawers need to be cleaned, hand-wash them with soap and water.These storage drawers are designed to be perfectly stacked on top of each other, but they can also be used individually based on your organizing needs. Each drawer has silicone pads on the bottom to ensure they won’t slip and move around when using them. They have smooth pull-out handles for quick access to your belongings. I actually prefer this stackable drawer design to my own set. The one flaw in my own plastic organizer was that the handles were placed at the top, which put pressure on the sides, causing one plastic drawer to snap when I didn’t support it from the bottom when I pulled it out. Since these storage drawers have the handles at the bottom, it better distributes the weight.Related: Walmart is selling a storage pantry cabinet for $120 that can hold up to 720 poundsWhy do shoppers love it?A major reason these storage drawers have become a bestseller is that they’re so popular among shoppers. “These are superb for storing all my makeup,” raved one shopper. “It keeps it organized yet clutter-free, which was my end goal! Quality of the build is really good, too!”Another reviewer wrote, “I think these are fabulous, very sturdy, and give my bathroom a stylish, clean look. I just have so much stuff, so glad I found these to help organize my mess!”Shop more dealsChoezon Stackable Closet Storage Drawers, $33 (was $60) at AmazonAmazon Basics Stackable Plastic Organizer Drawers, $20 (was $21) at AmazonVtopmart 44-Piece Clear Plastic Drawer Organizers Set, $24 (was $32) at AmazonDon’t miss your chance to score the Vtopmart 5-Piece Stackable Storage Drawers Set for only $44 at Amazon. This limited-time deal could soon end, so secure the savings now to step up your organization for less.

The U.S. economy is closer to a breaking point than it looks at this point. According to a Barrons’ report, Moody’s chief economist Mark Zandi has reset his recession odds to a worrying 49%, arguing the economy looked shaky well before the Iran-driven surge in oil prices.Layer that on top of remarkably weak jobs data and sluggish GDP growth, and he points to a credible risk that higher energy costs would tip the economy over the line.Though the Iran conflict isn’t the root cause of the problem, it is a potential trigger that could turn a soft patch into something a lot worse.As of Wednesday, March 18, U.S. WTI crude was around $94.50 to $94.65 a barrel, Reuters reported, and Brent was around $102.93 to $103.12, having surged from $67 a barrel before the Iran war began in late February. Also, U.S. gasoline prices have jumped by nearly 84 cents a gallon to over $3.75, their highest average since October 2023, according to Reuters. Interestingly, in my previous piece on Zandi, he spoke about the economy’s fragility beneath the surface.He pointed to relatively solid GDP growth, slowing job creation, and accelerating AI-driven productivity as the primary reasons for the sluggishness. If AI-led demand doesn’t rise quickly enough, the unemployment picture would start to look uglier. So clearly, he has been sensing the economy’s vulnerability for some time, and the latest jolt is arguably a lot more pronounced.Where Wall Street stands on recession odds right nowMohamed El-Erian: 35% — Bumped odds from 25% as oil prices rise, and consumer strength weakens.Goldman Sachs: 25% — Lifted recession risk following soft jobs data and growing geopolitical pressure.J.P. Morgan: 35% — Sees elevated recession risk for 2026, arguing that recovery remains mostly fragile.Apollo/Torsten Sløk: 30% — Market-implied odds are near 30%, still behind the more bearish calls.
Source: Barrons, Fortune, Business Insider, JP Morgan, Apollo

Moody’s economist Mark Zandi warns of a recession risk near the tipping point, amid an oil surge pressuring the economy.Williams/CQ-Roll Call via Getty Images

Who is Mark Zandi?Zandi is one of the most influential economists on Wall Street, sitting at the crossroads of markets, policy, and real-world predictions.Currently, he serves as the chief economist at Moody’s Analytics, directing economic research, and he co-founded Economy.com in 1990, which Moody’s acquired in 2005. Collectively, he has been in the game for nearly 35 years, making him a bona fide veteran that the big banks and lawmakers still listen to.MoreEconomic Analysis:Ernst & Young drops blunt reality check on the economyFederal Reserve official blasts latest interest-rate pauseIMF drops blunt warning on US economyMoreover, his calls have consistently been prescient and well-timed ahead of the market.Zandi rose to center stage during the financial crisis.In November 2008, he sounded the alarm in the Senate that the U.S. was likely to experience its worst economic setback since the Great Depression, making him one of the most prominent mainstream voices during the 2008 meltdown. On top of that, he boasts real Washington credibility. For context, Zandi served as an unpaid economic adviser to John McCain’s 2008 presidential campaign and has testified before Congress multiple times. The weak jobs report that raised fresh recession concernsThe latest U.S. jobs report has sparked intense discussion over the past few days, showing negative hiring while unemployment moved higher, sharpening fears that the labor market is losing momentum.Payrolls dropped by 92,000 in February, compared to market expectations for a 59,000 gain.Unemployment rose to 4.4% from 4.3%, which shows that labor slack is building.Core sector-wide job losses include health care, which saw a 37,000 drop in physicians, while information technology shed 11,000 jobs, and federal government employment fell by 10,000.Payrolls have now posted the sixth consecutive monthly drop since early 2025, with job gains averaging just 6,000 over the past three months.
Source: Reuters, Bureau of Labor Statistics
Zandi warns the U.S. economy may be one shock away from recessionZandi’s recession odds, noted by Barron’s, were just below a tipping point, even before the recent events in the Middle East.Now, the veteran economist says the real risk is that higher energy prices could push it over the edge, especially with labor and GDP growth already softening. The economy was already losing momentum: The February jobs report showed 92,000 jobs lost, with Zandi arguing that “almost all” economic data has turned soft since late last year. Growth has been sluggish, too, with GDP revised down to 0.7% in Q4, comfortably below prior estimates, and a steep drop from 4.4% in Q3.Oil is the potential tipping point: Zandi feels it wouldn’t take much for recession odds to cross 50%, noting that virtually every post-WWII downturn has been preceded by rising oil prices. Elevated energy costs squeeze demand and act like a tax to consumers and businesses.The risk is broader than recession alone: Another top economist, Nobel laureate Joseph Stiglitz, said the U.S. might face stagflation, led by choppy growth and persistent inflation. Not everyone is buying the oil-driven recession callThough plenty has been made of the Iran conflict’s impact on the economy, not every big-name analyst feels it’s a long-term demand killer.For instance, Goldman Sachs raised its Q4 oil forecasts, but its base case still assumes that the Strait of Hormuz flows start recovering with WTI moderating in the $70s by early June following fresh reserve releases and normalization in logistics. However, as I covered in a recent story, big-bank analysts also believe that if the conflict becomes more prolonged than expected, conditions could turn rocky.Moreover, Bank of America’s global fund manager survey underscores a similar narrative, Reuters suggests. Investors, despite concerns about stagflation, still expect Brent to settle around $76 by year-end, suggesting they see today’s oil spike as severe but not irreversible.Related: Veteran analyst sends blunt message on Nvidia stock after GTC

The five largest U.S. hyperscalers plan to spend roughly $660 billion to $690 billion on AI infrastructure in 2026 alone, according to Futurum Group research published in February. That figure nearly doubles what the same companies spent just 12 months earlier, per CreditSights projections.Nvidia stock is up, Broadcom stock is up, and every brokerage in America has an AI picks list ready for your screen. If you are a long-term investor, you might assume you have already heard everything worth hearing about this trade.You have not.A senior technology specialist at Wellington Management, one of the oldest active managers in the country, just laid out a thesis that reframes the entire AI opportunity. His argument does not center on which stock to buy next quarter; it centers on a shift that has barely begun.The AI ecosystem has layers that most investors overlookBrian Barbetta is a technology and AI specialist at Wellington Management and co-portfolio manager of the Vanguard Wellington U.S. Growth Active ETF (VUSG) and the Vanguard Global Equity Fund (VHGEX). In a Q&A published on Vanguard’s advisor insights page, Barbetta broke the AI sector into four distinct layers that matter for your portfolio decisions going forward.Four layers of the AI sector you should understandInfrastructure: Companies building the physical backbone, including data centers, power systems, and semiconductor makers like Nvidia and Broadcom.Enablers: Foundational large language model creators such as OpenAI and Anthropic, plus cloud providers such as Google, Microsoft, and AmazonApplications: Software companies like Adobe that embed AI into products you already use, including coding copilots and image generation toolsBeneficiaries: Health care providers, banks, and financial institutions using AI to improve efficiency across their existing operations and servicesMost investors focus only on the first layer, buying chipmakers and hoping for the best without understanding the full value chain. Barbetta pointed out that the hyperscalers now span multiple layers at once, which makes their competitive position uniquely durable.Reasoning models changed the AI game in a single yearIf you think AI peaked with the launch of ChatGPT in late 2022, Barbetta’s timeline suggests you are behind at least one major leap. He identified the introduction of reasoning models in 2025 as the single biggest evolution since ChatGPT’s debut shook global markets.The original large language models simulated the most statistically likely answer to your question, which made them impressive but unreliable. Reasoning models changed the process entirely by allowing the AI to critique and refine its own responses before delivering a final answer.These newer models require significantly more computing power, which explains the record-breaking infrastructure spending you keep reading about. Goldman Sachs projects that hyperscaler capital expenditure from 2025 through 2027 will reach $1.15 trillion, more than doubling the $477 billion spent over the prior three years.For your portfolio, this means the demand driving AI infrastructure stocks is not speculation about a distant future or wishful thinking. Real usage of these systems has grown exponentially, and the computing requirements behind that usage keep expanding in measurable ways.Agentic AI could be the real portfolio game-changerHere is where Barbetta’s thesis takes a turn that should get your attention if you invest in technology or use it daily. He described agentic AI as the next major unlock, a system that does not just answer your questions but completes tasks on your behalf.Imagine giving your AI assistant access to Excel, Outlook, Bloomberg, and the internet, then asking it to complete a full research project. That is not science fiction or marketing hype, according to Barbetta, though he cautioned it remains early to say whether 2026 delivers it.What agentic AI could mean for everyday investorsThe practical implications extend well beyond Wall Street trading desks and institutional research departments into your household decisions. Barbetta offered a simple consumer example that captures the scale of what could happen if agentic AI reaches mass adoption soon.Related: Galaxy S26 brings ‘agentic AI’ to phones, and it’s bigger than SamsungInstead of asking AI to plan your vacation, you could let it take control of your browser and actually book it for you completely. The same logic applies to managing your finances, scheduling appointments, comparing insurance quotes, and handling routine administrative tasks.For investors, the key insight is that preparing all data for AI today is labor-intensive and limits how useful these systems truly become. Letting AI interact directly with software programs could dramatically expand the total number of use cases and revenue streams available.Why this AI cycle looks different from the dot-com crashEvery time AI stocks rally, someone compares the moment to 1999, and you have probably heard the dot-com bubble analogy more than once. Barbetta pushed back on that comparison directly by pointing to one critical difference separating today’s environment from that earlier era.More AI Stocks:Morgan Stanley sets jaw-dropping Micron price target after eventBank of America updates Palantir stock forecast after private meetingMorgan Stanley drops eye-popping Broadcom price targetMost AI spending today shows positive, measurable return on invested capital across both infrastructure and the cloud, he told Vanguard. The infrastructure being built is producing attractive returns using reasonable financial assumptions, unlike the speculative dot-com buildout of the 1990s.Where the real AI bubble risk exists right nowBarbetta did not dismiss the AI bubble risk entirely, and you shouldn’t either if you are building a long-term portfolio around AI exposure. He specifically flagged pockets of overexcitement in private markets where some investors may not fully understand the risks they are taking.He also warned about neocloud companies that rent GPU computers to hyperscalers at inflated prices driven by the current supply shortage. When that supply eventually catches up with demand, these companies could face severe economic challenges that their current valuations do not reflect.A Moody’s Ratings report published in March 2026 noted the combined backlog of contracted revenue across key hyperscalers has reached approximately $1.7 trillion. That contracted demand offers real support, but it does not guarantee that every company riding the AI wave will survive the eventual shakeout.How Vanguard’s specialist identifies real AI winnersNot every AI stock that looks like a leader today will remain one, and Barbetta was blunt about what separates true winners from pretenders. His framework focuses on a single principle that should guide every investment decision you make inside this rapidly evolving sector right now.The moat test that separates winners from also-ransCompanies that compete primarily on price will always have limited returns, no matter how fast their revenue grows in the short term. True long-term winners build natural monopolies or near monopolies with clear competitive advantages and defensible moats around their businesses.Barbetta broke down what specific advantages he sees at the major players that could determine the next decade of AI market leadership.Google: Lowest cost to serve because of custom hardware, vast proprietary data, and enormous distribution through its applications and devicesMeta: Applies AI to advertising technology where financial returns on investment are among the highest of any business model worldwideOpenAI: Built enormous consumer usage with ChatGPT, though whether that lead holds as competitors like Google Gemini advance remains uncertainAnthropic: Has demonstrated exceptional engineering strength in code generation on Claude, which may give it a natural edge among enterprise developersMeanwhile, the hyperscalers share one durable advantage that smaller competitors cannot match, regardless of how clever their technology may be. They can redeploy massive amounts of capital continuously and keep growing at a scale that creates compounding benefits over long time horizons.Market concentration creates an opportunity for active investorsIf you own an S&P 500 index fund, you already have heavy exposure to the same mega-cap AI companies dominating headlines every single week. Barbetta argued that this concentration actually creates a unique opportunity for active managers willing to take calculated, benchmark-relative risks.His approach uses a benchmark-relative active risk framework that asks one clear question about every position inside the portfolio he manages. That question is straightforward but powerful: How much risk does each position carry relative to the benchmark, and what does it look like overall?

Not every AI stock that looks like a leader today will remain one, says Technology and AI Specialist at Wellington Management Brian Barbetta.d3sign/Getty Images

When a big bet against a company can generate alphaSignificant overweights and underweights are both useful tools in concentrated markets where a few stocks drive the majority of returns. If a company is on the wrong side of a major technological change, a large underweight position can generate meaningful alpha for the portfolio.Related: Cathie Wood buys $2 million of tumbling AI stockWhen the market overreacts to a negative headline, as Barbetta says it occasionally does with AI leaders like Google, active managers can lean in.Wellington-managed Vanguard funds have outperformed peer-group averages in 11 of 12 solely managed funds over the 10-year period ending September 2025.For your own portfolio, the lesson is that passive exposure alone may not capture the full opportunity in an AI-driven market. Consider whether a blend of index funds and actively managed strategies might help you navigate the winners and losers more effectively.How you can use AI to sharpen your own financial decisionsBarbetta’s advice was not limited to professional money managers or institutional advisors with Bloomberg terminals on their desks every morning. He made a direct recommendation that applies to any investor, financial advisor, or household trying to make smarter decisions with available tools.His core message was simple: People who embrace technology advancements early can quickly become super performers in their respective fields. Rather than fearing AI disruption, Barbetta suggested leaning in and learning exactly how these tools work to strengthen your financial life.Practical steps you can take right now:Load information about your investments, goals, and risk tolerance into your preferred AI model to generate personalized preparation memos.Use AI to retrieve and analyze financial data before making portfolio decisions, then go back and forth to refine your research approach.Let AI handle routine data tasks so you can focus your energy on the high-value, personal decisions that actually require human judgment.Start small by experimenting with free AI tools for budgeting, tax estimation, or retirement planning before committing to paid services.The biggest risk is not that AI replaces your judgment as an investor or a professional managing your own household financial decisions. The biggest risk, according to Barbetta, is that someone else masters these tools first and gains an advantage you cannot easily close later.What this means for your portfolio in 2026 and beyondThe AI trade is entering a new phase, and the usual strategy of buying chipmakers and hoping for the best will not be enough going forward. Vanguard’s global chief economist Joe Davis has said the firm expects AI to stand out among megatrends for its capacity to transform labor markets.Davis compared the current AI investment cycle to the railroad buildout of the mid-19th century and the 1990s telecommunications infrastructure surge. Vanguard projects up to a 60% chance that the U.S. economy achieves 3% real GDP growth in coming years, driven partly by AI investment.Your next step should not be chasing the latest AI stock tip or loading up on a single semiconductor name that dominated last quarter’s news. Instead, focus on understanding which companies in your portfolio have genuine moats, real returns on invested capital, and durable competitive edges.Agentic AI is coming, and it will reshape how you interact with your money, your advisor, and your daily life in ways that feel inevitable later. The investors who understand this shift before it becomes consensus will be the ones best positioned for what the next five years actually deliver.Related: Major tech CEO sounds alarm on AI agents

Many founders assume growth slows because of product or market issues. In reality, unclear and undisciplined communication often becomes the hidden bottleneck.

In the last few years, Chinese AI startup MiniMax has become one of the most exciting in the crowded global AI marketplace, carving out a reputation for delivering frontier-level large language models (LLMs) with open source licenses and before that, high-quality AI video generation models (Hailuo). The release of MiniMax M2.7 today — a new proprietary LLM designed to perform well powering AI agents and as the backend to third-party harnesses and tools like Claude Code, Kilo Code and OpenClaw — marks yet a new milestone: Rather than relying solely on human-led fine-tuning, MiniMax has leveraged M2.7 to build, monitor, and optimize its own reinforcement learning harnesses. This move toward recursive self-improvement signals a shift in the industry: a future where the models we use are as much the architects of their progress as they are the products of human research. The model is categorized as a reasoning-only text model that delivers intelligence comparable to other leading systems while maintaining significantly higher cost efficiency.However, with M2.7 being proprietary for now, it is a sign once again that Chinese AI startups — for much of the last year, the standard-bearers in the world of the open source AI frontier, making them appealing for enterprises globally due to low (or no) costs and customization — are shifting strategy and pursuing more proprietary frontier models like U.S. leaders like OpenAI, Google, and Anthropic have been doing for years. MiniMax becomes the second Chinese startup to release a proprietary cutting-edge LLM in recent months following z.ai with its GLM-5 Turbo, and rumors that Alibaba’s Qwen team is also shifting to proprietary development in the wake of the departure of senior leadership and other researchers.Technical achievement: The self-evolution loopThe defining characteristic of MiniMax M2.7 is its role in its own creation. According to company documentation, earlier versions of the model were used to build a research agent harness capable of managing data pipelines, training environments, and evaluation infrastructure. By autonomously triggering log-reading, debugging, and metric analysis, M2.7 handled between 30 percent and 50 percent of its own development workflow. This is not merely an automation of rote tasks; the model optimized its own programming performance by analyzing failure trajectories and planning code modifications over iterative loops of 100 rounds or more.”We intentionally trained the model to be better at planning and at clarifying requirements with the user,” explained MiniMax Head of Engineering Skyler Miao on the social network X. “Next step is a more complex user simulator to push this even further.”This capability extends to complex environments via the MLE Bench Lite, a series of machine learning competitions designed to test autonomous research skills. In these trials, M2.7 achieved a medal rate of 66.6 percent, a performance level that ties with Google’s new Gemini 3.1 and approaches the current state-of-the-art benchmarks set by Anthropic’s Claude Opus 4.6. The goal, according to MiniMax, is a transition toward full autonomy in model training and inference architecture without human involvement. Performance evolution: MiniMax m2.7 vs. m2.5When compared to its predecessor, M2.5, released in February 2026, the M2.7 model demonstrates significant gains in high-stakes software engineering and professional office tasks. While M2.5 was celebrated for polyglot code mastery, M2.7 is designed for real-world engineering—tasks requiring causal reasoning within live production systems.Key performance metrics include:Software engineering: M2.7 scored 56.22 percent on the SWE-Pro benchmark, matching the highest levels of global competitors like GPT-5.3-Codex.Professional office delivery: In document processing, M2.7 achieved an Elo score of 1495 on GDPval-AA, which the company claims is the highest among open-source-accessible models.Hallucination reduction: The model scores plus one on the AA-Omniscience Index, a massive leap from the negative 40 score held by M2.5.Hallucination rate: M2.7 achieves a hallucination rate of 34 percent, which is lower than the rates of 46 percent for Claude Sonnet 4.6 and 50 percent for Gemini 3.1 Pro Preview.System comprehension: On Terminal Bench 2, the model scored 57.0 percent, demonstrating a deep understanding of complex operational logic rather than simple code generation.Skill adherence: On the MM Claw evaluation, which tests 40 complex skills exceeding 2,000 tokens each, M2.7 maintained a 97 percent adherence rate, a substantial improvement over the M2.5 baseline.Intelligence parity: The model’s reasoning capabilities are considered equivalent to GLM-5, yet it uses 20 percent fewer output tokens to achieve similar results.The model’s evolution is further evidenced by its score of 50 on the Artificial Analysis Intelligence Index, representing an 8-point improvement over its predecessor in just one month, and also taking the 8th place overall globally in terms of its overall intelligence across benchmarking tasks in various domains.Not all independent, third-party benchmarks show improvement for M2.7 over M2.5: On BridgeBench, a set of tasks designed by agentic AI coding startup BridgeMind to test a model’s performance for “vibe coding,” or turning natural language into working code, M2.5 scored 12th place while M2.7 scored 19th place.Access, pricing, and integrationMiniMax M2.7 is a proprietary model available through the MiniMax API and MiniMax Agent creation platforms. While the core model weights for M2.7 remain closed, the company continues to contribute to the ecosystem through the open-source interactive project OpenRoom. For direct API integration and via third-party provider OpenRouter, MiniMax M2.7 maintains a cost-leading price point of 0.30 dollars per 1 million input tokens and 1.20 dollars per 1 million output tokens, which is unchanged from the pricing for M2.5.To support different usage scales and modalities, MiniMax offers a structured Token Plan with various subscription tiers. These plans allow users to access models across text, speech, video, image, and music under a single unified quota. To further drive adoption, MiniMax has launched an Invite and Earn referral program, providing a 10 percent discount to new invitees and a 10 percent rebate voucher to the inviter.Monthly standard Token Plan pricing: The standard monthly tiers are designed for entry-level developers to heavy regular users.Starter: $10 per month for 1,500 requests per 5 hours.Plus: $20 per month for 4,500 requests per 5 hours.Max: $50 per month for 15,000 requests per 5 hours.Monthly high-speed Token Plan pricing: For production-scale workloads requiring the M2.7-highspeed variant, the following tiers are available:Plus-Highspeed: $40 per month for 4,500 requests per 5 hours.Max-Highspeed: $80 per month for 15,000 requests per 5 hours.Ultra-High-Speed: $150 per month for 30,000 requests per 5 hours.Yearly Token Plan pricing: Yearly subscriptions provide significant discounts for long-term commitment:Standard Starter: $100 per year (saves 20 dollars).Standard Plus: $200 per year (saves 40 dollars).Standard Max: $500 per year (saves 100 dollars).High-Speed Plus: $400 per year (saves 80 dollars).High-Speed Max: $800 per year (saves 160 dollars).High-Speed Ultra: $1,500 per year (saves 300 dollars).One request in these plans is roughly equivalent to one call to MiniMax M2.7, though other models in the suite, such as video or high-definition speech, consume requests at a higher rate.Official tool integrationsTo ensure seamless adoption, MiniMax has provided official documentation for integrating M2.7 into over 11 major developer tools and agent harnesses. This includes widely used platforms such as Claude Code, Cursor, Trae, and Zed. Other officially supported tools include OpenCode, Kilo Code, Cline, Roo Code, Droid, Grok CLI, and Codex CLI.Additionally, the model supports the Model Context Protocol, allowing it to natively use tools like Web Search and Understand Image for multimodal reasoning. Developers using the Anthropic SDK can easily integrate M2.7 by modifying the ANTHROPIC_BASE_URL to point to the MiniMax endpoint. When using MiniMax as a provider in tools like OpenClaw, image understanding capabilities are automatically configured via the model’s VLM API endpoint, requiring no extra setup from the user.With its deep bench of integrations and its pioneering approach to recursive self-evolution, MiniMax M2.7 represents a significant step toward an AI-native future where models are as involved in their own progress as the humans who guide them.Strategic implications for enterprise decision-makersTechnical decision-makers should interpret the M2.7 release as evidence that agentic AI has moved from theoretical prototyping to production-ready utility. The model’s ability to reduce recovery time for live production incidents to under three minutes by autonomously correlating monitoring metrics with code repositories suggests a paradigm shift for SRE and DevOps teams.Enterprises currently facing pressure to adopt AI-driven efficiencies must decide whether they are content with AI as a sophisticated assistant or if they are ready to integrate native agent teams capable of end-to-end full project delivery.From a financial perspective, M2.7 represents a significant breakthrough in cost efficiency for high-level reasoning. Analysis indicates that M2.7 costs less than one-third as much to run as GLM-5 at equivalent intelligence levels. For example, running a standard intelligence index cost 176 dollars on M2.7 compared to 547 dollars for GLM-5 and 371 dollars for Kimi K2.5. This aggressive pricing strategy places M2.7 on the Pareto frontier of the intelligence vs. cost chart, offering enterprise-level reasoning at a fraction of the market rate.The current market is saturated with high-performance models, many of which still hold slight edges in general reasoning scores. But the specific optimization of M2.7 for Office Suite fidelity in Excel, PPT, and Word and its high performance in the GDPval-AA benchmark make it a primary candidate for organizations focused on professional document workflows and financial modeling. Decision-makers must weigh the benefits of a general-purpose frontier model against a specialized engine like M2.7, which is built to interact with complex internal scaffolds and toolsets.Ultimately, the fact that it is fielded by a Chinese company (headquartered in Shanghai) and subject to that country’s laws in addition to the user’s country, and is not available for offline or local usage yet, may make it a tough sell for enterprises operating in the U.S. and the West — especially those in highly-regulated or government-facing industries. Nonetheless, the shift toward self-evolving models suggests that the ROI of AI investment will increasingly be tied to the recursive gains of the system itself. Organizations that adopt models capable of improving their own harnesses may find themselves on a faster iteration curve than those relying on static, human-only refinement. With MiniMax’s aggressive integration into the modern developer stack, the barrier to testing these autonomous workflows has dropped significantly, placing pressure on competitors to deliver similar native agent capabilities.