Tech
Thinking Machines amps up its bet against one-size-fits-all AI with its first open model, Inkling
Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, released its first in-house AI model Wednesday morning, called Inkling. And unlike the flagship models from OpenAI, Anthropic, or Google, it’s open-weight, meaning outside developers and companies can download it and modify it directly.
Inkling is a mixture-of-experts system with 975 billion total parameters, though it only draws on a fraction of that — about 41 billion — for any given task, a common design that keeps very large models faster and cheaper to run. It was trained on 45 trillion tokens of text, image, audio, and video, and reasons natively across all four, according to the company’s own release materials. For now, though, its outputs are limited to text, including code, styled artifacts, and structured data.
The model is Thinking Machines Labs’ first public proof point after a year and a half spent building AI infrastructure largely out of public view. Some of that work had already surfaced in a May research preview of “interaction models” — AI designed to listen and speak (and even interrupt) instead of stop and wait as with typical chatbots. It’s also a test of the central bet behind the startup, which is that AI that organizations can adapt for themselves will outperform the one-size-fits-all models the biggest labs currently sell.
Inkling is designed to give calibrated answers, including flagging uncertainty rather than guessing, and lets users dial “thinking effort” up or down when they want to trade for speed. On one benchmark, the company says, Inkling uses a third as many tokens as Nvidia’s Nemotron 3 Ultra — its latest generation open-weight model — to hit the same coding performance.
Thinking Machines doesn’t claim Inkling is best-in-class. Its newest blog post states explicitly that Inkling is “not the strongest overall model available today, open or closed.” What it’s evidently going for instead is well-rounded performance.
That raises the question of who, within the enterprise market it’s targeting, this product is really for. Thinking Machines is, for now, marketing Inkling less as a finished product than as a starting point, something for organizations to fine-tune themselves through Tinker, the company’s model-customization platform. This also means customers, not Thinking Machines, are responsible for making sure their customizations are safe, for example. (Fine-tuning requires serious machine-earning talent.)
OpenAI, Anthropic, and Google have all taken a very different approach with ChatGPT, Claude, and Gemini, respectively, which were all built to compete as general-purpose chatbots first, with agentic, autonomous features layered on top.
A post published by Thinking Machines last week was clearly meant as the backdrop for this release. AI that’s trained centrally by one company and then set in stone, the company argued in that post, underperforms AI that organizations shape themselves because so much expertise is specific to the people who hold it.
Other arguments against closed models are gaining steam. In a blog post published Sunday, Microsoft CEO Satya Nadella — whose company has invested billions in both OpenAI and Anthropic — warned that enterprises using proprietary AI models effectively pay twice: once in subscription costs, and again by handing over business knowledge embedded in their prompts and corrections, which can be absorbed into future model versions.
Hugging Face CEO Clem Delangue made a similar prediction in conversation with TechCrunch last week. Frontier models, he said, will increasingly be reserved for experimentation and high-value tasks, while most production AI work shifts to private or open-source alternatives — the exact split Thinking Machines is building around.
The clearest argument for Thinking Machines’ approach came from a recent project with Bridgewater Associates, the world’s largest hedge fund (which is not, for what it’s worth, a Thinking Machines investor). Researchers from both companies took an existing open-source model and trained it further on Bridgewater’s own financial expertise. The result was said to score 84.7% on financial reasoning tests, beating top proprietary AI models, while costing roughly a fourteenth as much to run — though those results come from the two companies’ own evaluation, not an independent one.
Either way, Thinking Machines is emphasizing how quickly it got here. OpenAI took roughly five years to bring its tech to market and show revenue, and Anthropic roughly three. Thinking Machines says it did the same in about nine months.
Some will wonder whether Inkling was trained on outputs from competitors’ models, a practice known as “distillation” that has drawn scrutiny across the industry. The short answer, per the company’s own materials, is partly. Thinking Machines pre-trained Inkling from scratch, but it says it used other open-weight models — including Moonshot AI’s Kimi K2.5 — to help generate some of its early post-training data before large-scale reinforcement learning took over. The next model, the company insists, will use fully self-contained post-training instead.
On the cost side, Thinking Machines has been more guarded. It struck a partnership with Nvidia in March to deploy a gigawatt of Vera Rubin computing capacity and trained Inkling entirely on Nvidia’s GB300 NVL72 systems — but hasn’t said how it plans to cover those costs, and revenue, by most accounts, hasn’t been a priority. (A reported $50 billion fundraising round was said to be coming together last November but had stalled by January; the company has declined to talk about its funding picture since.)
A related question is whether Thinking Machines’ spending will ever reach the scale of OpenAI’s or Anthropic’s, or whether its efficiency-driven approach means the economics look different. Put another way, the company’s bet may be less that it will eventually spend like its larger rivals than that it won’t need to at all — because once weights are public, nothing obligates anyone who downloads them to pay Thinking Machines to run them, unlike the metered access OpenAI and Anthropic sell. It’s Tinker, not the model itself, where the company’s revenue has to come from, via training, fine-tuning, and, now, a cut of the hosting ecosystem built around it.
Headcount, at least, looks more settled. Thinking Machines now employs roughly 200 people, up from levels reported after a wave of departures earlier this year, including two co-founders who left for OpenAI in January.
Thinking Machines, for its part, doesn’t seem interested in playing up individual moves the way much of the industry does. According to a source inside the company, its culture, by design, favors continuity over reliance on any one personality. It makes sense: it’s less of a setback when people change teams if they were never put on a pedestal to begin with. It’s also a remarkable thing for a company to insist on, given how much of its own story is still associated with the name of its now-famous co-founder, whether she planned it or not.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
Federal employees can download TikTok on their work phones again
The Department of Justice says that federal employees can now download TikTok on their government devices, according to Reuters.
A 2022 law banned federal employees from using the short-form video app on those devices, but the DOJ reportedly says the law no longer applies, thanks to a deal transferring ownership of TikTok’s U.S. operations to a joint venture backed by Oracle, Silver Lake, and MGX. (Oracle serves as the security partner for the new joint venture, while previous owner ByteDance retains a 19.9% stake.)
The DOJ memo reportedly says President Donald Trump has cleared “employees of Executive Branch agencies” to “download TikTok onto their official devices, subject to the agency’s discretion and consistent with all applicable workplace policies.”
Following the ban focused on government employees and devices, the app was banned more broadly across the United States. But just as the law took effect early last year, the app only went down briefly before Trump repeatedly delayed the move and urged service providers to restore access.
Tech
All the EVs that were discontinued or killed off in the U.S. this year
The Honda Prologue, you may have heard, is officially dead — a decision the company confirmed to TechCrunch, removing the last all-electric vehicle from the automaker’s U.S. portfolio. The Prologue’s departure signals more than Honda’s EV backpedaling. It also illustrates a broader EV industry retreat from the U.S. market (in stark contrast to the rest of the world).
The demise of the Honda Prologue got us thinking: What other EVs have left the U.S., and why?
The end of the $7,500 federal tax credit had an outsized effect on EV sales in the United States. But there are other reasons behind the winnowing choices, including tariffs, changing consumer tastes, costs, company priorities, and regulatory action. According to data published in July by Kelley Blue Book and Cox Automotive, 247,226 EVs were sold in the second quarter or about 5.8% of the total market. While EV sales grew between the first and second quarters of 2026, they are still down from the same period last year (and before that tax credit ended in fall 2025).
Still Americans are still buying EVs, and there are new EVs entering the U.S. market — the Rivian R2 is one example. And there are signs of a slow recovery. Fourth quarter 2025 sales were 36% lower than the same period in 2024. This year that gap has narrowed, albeit still below sales figures from the previous year. For example, EV sales in Q2 were 20.5% lower than the same period in 2025.
Even with a recovery underway, automakers are pulling the plug on many EV modes. Here are those ones that have left or are leaving. TechCrunch will periodically update this list of EVs that have left, or are leaving, the U.S. market in 2026.
Afeela
Ah, Afeela we never even knew ya.
The Afeela got its start as the Vision S, a prototype announced by Sony in 2020 at the Consumer Electronics and that ended up being one of the big, surprising reveals of the annual tech trade show. Honda entered the picture in 2022 when the two Japanese conglomerates announced a joint venture; they showed off an Afeela-branded prototype the following year.
In the months and years that followed, there was constant barrage of updates about the Afeela, which seemed to be everywhere, and yet nowhere. It was even displayed at TechCrunch Disrupt one year.
The Afeela, despite the marketing blitz, never made it into production. In March 2026, the joint venture gave up on the two Afeela-branded EVs. The move followed Honda’s decision, announced just a two weeks before, to cancel three EVs planned for the U.S. market.
Honda (and Acura!)
It was just a couple of years ago that Honda declared its EV ambitions with its O Series, including a mid-sized SUV prototype that debuted at the CES 2025 tech trade show and its futuristic Saloon and Space-Hub concepts the year before. The SUV, which was slated for production at Honda’s “EV Hub” factory in Ohio, was supposed to debut in North America in the first half of 2026.
Honda stopped development of the Acura RDX, Honda O sedan and SUV in March 2026 as part of a major overhaul of the company’s EV plans. The company blamed U.S. tariffs and Chinese competition for the decision.
There was also chatter at the time that Honda was planning to stop production of the Prologue, but there was no official announcement until July 16 when CarBuzz was the first to report that the Prologue program was ending. TechCrunch confirmed with Honda that the Prologue was going out of production.
The death of the Series 0 is difficult to measure since it never went into production. The Prologue represented more grounded goals than the O Series, and one that actually went into production and sold to U.S. consumers. The Prologue was a product of a partnership with General Motors — it is built at GM’s Ramos Assembly Plant in Mexico — and closely related to the Chevrolet Blazer EV. And it did OK for awhile, selling roughly 33,000 units in 2024 and 39,000 in 2025, before the tax credit ended and sales went into a free fall.
Hyundai
The Korean automaker has actually done quite well selling EVs to Americans. But it has made a few changes based on changing economics. In March, the company said it would no longer sell the Hyundai Ioniq 6 in the U.S., a decision that was likely tied to tariffs. The Ioniq 6 is made in South Korean and imported to the U.S., while its Ioniq 5 and Ioniq 9 models are assembled at its Georgia factory.
The company has said it will continue to import its more expensive, lower volume N-model of the Ioniq 6.
Nissan
Nissan decided last year it would not produce a 2026 model year of its all-electric Ariya SUV for the U.S. market. And it doesn’t appear to be returning. Nissan first unveiled the Ariya in 2020 and planned to start selling it in Japan the following year.
The Ariya was the first all-electric to come out of Nissan since the early EV pioneer introduced the Leaf hatchback a decade ago.
Polestar
Swedish EV maker Polestar, owned by Chinese automotive giant Geely, has been forced to leave U.S. over the country’s ban on Chinese-connected vehicle technology. Polestar needed specific authorization from the U.S. Department of Commerce to continue importing and selling its vehicles in the United States.
Without it, Polestar has been effectively banned from the United States. The company said it would continue selling its existing stock of Polestar 3 and Polestar 4 vehicles in the U.S., and that it will “continue to support customers, including providing access to its service network.” The Polestar 3 was assembled at a factory in South Carolina and in Chengdu, China.
Volvo Cars, Polestar’s sibling company that is also owned by Geely, did receive the authorization.
Tesla Model S and Tesla Model X
Tesla announced in January that it would end production of the Model S sedan and Model X SUV to make way for what the company views is the future. And it’s not a traditional electric sedan or SUV. In Tesla’s view, the future is AI, autonomy, and robots. It’s worth noting that sales of the S and X have fallen steadily over the years as consumers turned to its high volume and cheaper vehicles, the Model 3 and Model Y.
The last Model S and X vehicles rolled off the assembly line this spring. The company recently removed the assembly lines for the S and X at its Fremont, California factory to make room for production of its Optimus robots.
Volkswagen
Volkswagen has pulled back on the ID. 4 electric SUV and the ID Buzz.
In April, Volkswagen said it would no longer produce the ID.4 at its U.S. factory in Chattanooga, Tennessee in a shift to high-volume vehicles like its upcoming gas-powered Atlas SUV. The company said, at the time, U.S. customers will be able to buy the ID.4 until the current inventory runs out. VW said it expects U.S. inventory to last into 2027.
To be clear, Volkswagen has said the ID Buzz is merely on a hiatus and will return in 2027. But there is no 2026 model.
There are, however self-driving versions of the ID buzz currently being tested in the United States. Volkswagen subsidiary MOIA America and Uber started testing autonomous microbuses in Los Angeles in April in preparation for a robotaxi service that is supposed to launch in late 2026. When the service initially launches there the vehicles will have himan safety operators.
Volvo
Volvo decided in March that it would pull its subcompact EX30 and EX30 Cross Country variant from the U.S. market. The company said at the time that production for the U.S. would end sfter the summer. The EX30 had a promising start. It recieved a lot of attention prior to it official entry into the U.S. in 2025, and it was the company’s more affordable EV option.
Volvo does plan to continue selling the larger, all-electric EX60 and EX90 SUVs in the United States.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Tech
Kimi: Threat or menace?
Chinese company Moonshot AI released a new version of its Kimi model this week, leading to a perhaps-inevitable wave of discourse about China and open source AI.
Moonshot said that although Kimi K3 “still trails the most powerful proprietary models, Claude Fable 5 and GPT 5.6 Sol,” the new open source model “demonstrated frontier-level performance across our evaluation suite, consistently outperforming other tested models.” Independent analyses from Arena.ai and Vals AI also suggested that Kimi is competitive with flagship frontier models.
The announcement, which coincided with a speech from Chinese president Xi Jinping at the World AI Conference in Shanghai, seems to have spooked Wall Street, with the Nasdaq dropping about 1% on Friday as investors sold off stocks in chip companies like Nvidia.
Many of the resulting posts from tech industry figures will sound familiar to those who remember the debate after another Chinese company, DeepSeek, released its open source R1 model in January 2025. Except now, everything seems heightened after the Trump administration’s tariff war with China, repeated fights over the national security threat supposedly posed by Anthropic, and as major AI companies prepare to finally go public.
For example, David Sacks — the Trump administration’s former AI czar and now co-chair of the President’s Council of Advisors on Science and Technology — contrasted Kimi’s progress with a United States that is “tying itself in knots: politicians and bureaucrats are banning new data centers, piling on state regulations, and pushing for new federal agencies to pre-approve frontier models. This is how you lose the AI race.” (The news also gave him an excuse to take a dig at Anthropic, calling Claude an example of “woke lobotomized models” that are “the enemy American competitiveness.”)
And former Uber CEO Travis Kalanick echoed complaints that Chinese are “distilling off” (i.e., being trained on the outputs of) American AI models.
“If distillation isn’t enforced against, then everyone should be able to distill from everyone else.. otherwise one arm [would be] tied behind American models’ backs,” Kalanick wrote. (Of course, American models have also been built on top of Chinese ones, specifically Kimi.)
Meanwhile, OpenAI’s head of strategic futures Dean Ball said that Kimi is “a very good model” whose performance probably can’t be “explained away by distillation or anything like that,” adding that he’s “personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks.”
In fact, Ball suggested that “probable outcome of an open-weight-model-dominant world is full AI communism,” where AI is treated as “a ‘public good’ which will ultimately be provided by the state as a kind of ‘digital public infrastructure.’”
“This future strikes me as a dystopian hellscape, but I’ve never met an open-weight models advocate who doesn’t ultimately concede this is where things end,” said Ball. He even suggested that the Trump administration (which he used to work for) will eventually realize it needs to “create large amounts of regulatory risk around the use of open-weight Chinese models.”
“You don’t need to ‘ban open source’ (one of the dumber motifs of AI policy discussion),” Ball said. “You just need to direct every agency to issue soft law that creates FUD [fear, uncertainty, and doubt]. ‘A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models.’ It needn’t be that well justified. You just create enough regulatory risk that every regulated enterprise backs off.”
However, Shakeel Hashim, editor of the AI-focused publication Transformer, argued that much of the worry is overblown, both because Kimi “likely does not have dangerous cyber capabilities,” and because the Chinese government will face “extremely similar incentives” to restrict open Chinese models once they develop those capabilities.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.
Rain threatens rematch between Yankees, Dodgers
Andres Chaparro, Nats try to top gaudy numbers vs. skidding A's
