All posts tagged: model

Samsung’s new Ultra foldable is impressive, but I’d pay close attention to this model

Samsung’s new Ultra foldable is impressive, but I’d pay close attention to this model

Published by skeptic

Kerry Wan/ZDNET Follow ZDNET: Add us as a preferred source on Google. ZDNET’s key takeaways Samsung launched three foldable phones at its Unpacked event in London. Its first ‘Ultra’ foldable, the Samsung Galaxy Z Fold 8 Ultra, is an impressive device.  However, the Samsung Galaxy Z Fold 8 has a unique form factor that’s very compelling. Now that Samsung Unpacked is over and all the phones have been announced, you may be clamoring to get your preorder in for the all-new Samsung Galaxy Z Fold 8 Ultra. It is, after all, the first Samsung foldable to bear the “Ultra” moniker.  Also: Everything announced at Samsung Unpacked 2026: New foldables, smartwatches, more But you might want to hold on for a moment and check out the Samsung Galaxy Fold 8, too. Yes, the Fold 8 Ultra has a lot going for it, but the Fold 8 makes a strong argument for itself, as well. Part of that is because Samsung knows what’s coming down the pipeline: a foldable iPhone — a product that could disrupt the …

Samsung Galaxy Z Fold 8 Ultra vs. Z Fold 7: How the first ‘Ultra’ foldable compares to last year’s model

Samsung Galaxy Z Fold 8 Ultra vs. Z Fold 7: How the first ‘Ultra’ foldable compares to last year’s model

Published by skeptic

Kerry Wan/ZDNET Follow ZDNET: Add us as a preferred source on Google. The hot foldable summer is here, and Samsung has finally joined the fray with its latest Galaxy Z Fold 8 Ultra, Z Fold 8, and Z Flip 8. While there are notable reasons to consider the latter two if you’re queuing up for an upgrade, it’s the Z Fold 8 Ultra model that should have any power user’s utmost attention. Also: Samsung Galaxy vs. Google Pixel: I’ve used both brands extensively, and this one’s better Putting the fancy branding aside, the Z Fold 8 Ultra is the direct successor to last year’s Z Fold 7, with a similar design, screen aspect ratio, and general features. But Samsung’s also addressed some key concerns this year, from the battery life and charging of its most premium phone to software features that truly leverage the shape-shifting form factor. Ahead of Unpacked, I spent an intimate morning with the Z Fold 8 Ultra, comparing it to the Z Fold 7, which has been in my pocket since last July. …

Musk says Tesla to ‘open source’ Model S and X, but prior release makes us skeptical

Musk says Tesla to ‘open source’ Model S and X, but prior release makes us skeptical

Published by skeptic

Elon Musk says Tesla plans to “open source” the Model S and Model X, and he’s pointing to the original Roadster as the template. The problem is Tesla’s Roadster release amounted to eight files on GitHub plus five document packages behind a login, with no open-source license attached to any of it. What Tesla actually put out for the Roadster In November 2023, Musk posted that “all design & engineering of the original Tesla Roadster is now fully open source,” and added, “Whatever we have, you now have.” We covered the release at the time. The full inventory is short. The teslamotors/roadster repo holds five CAN database (.dbc) files covering the Roadster 1.5 and 2.0 ESS and PEM buses, about 339 KB in total, plus a 309 MB diagnostic software ISO and two PDF quick-guides. That’s everything, aside from a README and a .gitattributes file. Advertisement – scroll for more content The commit history is worth a look. The DBC files went up on December 2, 2022, nearly a year before Musk announced any of …

Anthropic launches Claude Opus 5, a cheaper AI model for coding, agents and enterprise workflows

Anthropic launches Claude Opus 5, a cheaper AI model for coding, agents and enterprise workflows

Published by skeptic

Anthropic released Claude Opus 5 on Friday, a model the company says delivers nearly all the intelligence of its top-of-the-line Claude Fable 5 at half the cost — a launch that signals how the AI race is shifting from raw capability to the economics of daily use. The model, available immediately on all of Anthropic’s platforms, is priced at $5 per million input tokens and $25 per million output tokens, unchanged from its predecessor, Opus 4.8. It becomes the new default model on Claude Max, Anthropic’s premium consumer tier, and the strongest model available on Claude Pro. The positioning is deliberate. Anthropic is not claiming Opus 5 is its smartest model — that distinction still belongs to Fable 5, and rival systems retain an edge in certain domains. Instead, the company is making a subtler argument that may matter more to enterprise buyers: that the most economically important AI work happens in a middle band of difficulty, where near-frontier intelligence delivered efficiently and cheaply beats frontier intelligence delivered expensively. “Opus 5 as your daily driver, …

Mirror publisher to pay ‘substantial’ hacking damages to model Paul Sculfor

Mirror publisher to pay ‘substantial’ hacking damages to model Paul Sculfor

Published by Sam Flores

Paul Sculfor. Distributed by Hamlins The publisher of the Mirror will pay model Paul Sculfor “substantial damages” over allegations of phone hacking while he was dating actresses Jennifer Aniston and Cameron Diaz. Sculfor’s claim related to nine articles containing his private information published by Daily Mirror, Sunday Mirror and The People owner Mirror Group Newspapers between 2007 and 2009. It said the information could only have been obtained through unlawful information gathering, including as a result of intercepting voicemail messages relating to his relationships with Aniston and Diaz. Sculfor’s claim was the only one of five test cases against MGN allowed to continue following a preliminary trial this year. Other claimants were deemed to have brought their cases too late because they could have known earlier and were now outside the six-year time limit under the Limitation Act 1980. However High Court judge Mr Justice Fancourt said Sculfor was misled into believing his family and friends were to blame for the articles. The judge found he had been “partly but causally misled by MGN’s publications: …

Kimi K3 Explained: 3 Trillion Parameter Open Model

Kimi K3 Explained: 3 Trillion Parameter Open Model

Published by skeptic

The Kimi K3, as explored by Prompt Engineering, represents a streamlined yet capable sibling to the larger Kimi K3 model. While it operates with fewer parameters than its 3-trillion-parameter counterpart, the K3 focuses on delivering efficiency and accessibility for organizations with more constrained computational resources. Its design emphasizes a balance between performance and practicality, particularly excelling in scenarios like mid-scale coding projects and targeted enterprise applications. By using a reduced but optimized architecture, the Kimi K3 offers a compelling alternative for businesses seeking specialized AI solutions without the extensive hardware demands of larger models. In this overview, you’ll gain insight into how the Kimi K3 carves out its niche in the AI landscape. Explore its strengths in handling domain-specific tasks, its integration of multimodal capabilities and the trade-offs it makes to achieve cost-effectiveness. Additionally, learn about the practical considerations for deploying the K3, including its hardware requirements and operational scope. This breakdown provides a clear understanding of how the Kimi K3 fits into the broader AI ecosystem and what it offers to organizations navigating the …

“He Was Very Close To Being Arrested”: Epstein’s Paris Model Scout Found Dead At Home, Weeks After CNN Appearance

“He Was Very Close To Being Arrested”: Epstein’s Paris Model Scout Found Dead At Home, Weeks After CNN Appearance

Published by Jenni Sidey

Daniel Siad, the 69-year-old Paris modeling scout whose name appears nearly 2,000 times in the DOJ’s Epstein files, was found dead at his home in Colombes, northwest of Paris, on Monday. The deputy public prosecutor at Nanterre, Marie-Celine Lawrysz, confirmed the death Wednesday and said an investigation into the cause was opened that evening, with an autopsy to follow. Which is to say: officially, nobody knows anything yet. Siad’s lawyer told Reuters that “Daniel Siad never stopped proclaiming his innocence” – and that her client died of a heart attack. To AFP, she was more careful, saying that if it was a heart attack, the strain and anxiety of the case will have played its part. The autopsy, presumably, will referee. And if that first statement sounds familiar, it should: when Jean-Luc Brunel was found dead in 2022, his lawyers announced that “Jean-Luc Brunel never stopped declaring his innocence.” The French defense bar evidently keeps the line on file. For over a decade, per the document dumps Congress pried out of the DOJ under the …

Google’s Flagship Still Can’t Ship – So It Launched Token Austerity And A Hacking Model Only Governments Can Use

Google’s Flagship Still Can’t Ship – So It Launched Token Austerity And A Hacking Model Only Governments Can Use

Published by Jenni Sidey

Alphabet reports second-quarter earnings after today’s close – the first hyperscaler print since cheap Chinese tokens knocked the semiconductor index into a bear market. So naturally, Google chose the eve of that report to ship three new AI models, none of which is the one it promised. Getty Images The Tuesday launch consisted of Gemini 3.6 Flash, a cheaper workhorse whose headline feature is that it consumes fewer tokens; Gemini 3.5 Flash-Lite, a high-throughput model built for volume; and Gemini 3.5 Flash Cyber, a vulnerability-hunting model that ordinary users are not permitted to touch. Conspicuously absent: Gemini 3.5 Pro, the flagship Google unveiled at I/O in May with a promised June launch, which has now missed multiple targets. Then there is Gemini 3.5 Flash Cyber, which Google says achieves top-tier performance at finding, verifying, and patching software vulnerabilities inside its CodeMender agent – and which will be available exclusively to governments and vetted partners through a limited-access pilot, on account of what the company calls the technology’s dual-use nature. Which is of course aimed at …

Poolside drops Laguna S 2.1, an open-weight coding model that beats rivals 10x its size

Poolside drops Laguna S 2.1, an open-weight coding model that beats rivals 10x its size

Published by skeptic

Poolside, the San Francisco AI lab that has spent most of its three-year existence quietly selling coding models to governments and defense agencies, released its most capable model to date on Tuesday — and made an unusually aggressive bet that radical transparency, not raw scale, is how a smaller lab competes at the frontier. The model, Laguna S 2.1, is a 118-billion-parameter Mixture-of-Experts (MoE) system that activates only 8 billion parameters per token, supports a context window of up to 1 million tokens, and — according to benchmarks published by the company — matches or beats open models several times its size on agentic coding tasks. The weights are available immediately on Hugging Face under the permissive OpenMDW-1.1 license. The headline numbers are striking for a model this small. Poolside reports that Laguna S 2.1 scores 70.2% on Terminal-Bench 2.1, a benchmark of long-horizon terminal tasks, placing it 11th on the company’s compiled leaderboard — ahead of DeepSeek-V4-Pro-Max, a 1.6-trillion-parameter model that scored 64.0; Thinking Machines’ 975-billion-parameter Inkling, at 63.8; and Nvidia’s 550-billion-parameter Nemotron 3 …

Google’s Gemini 3.6 Flash model cuts AI agent token costs by up to 65% on long horizon engineering tasks —and 3.5 Pro is on the way

Google’s Gemini 3.6 Flash model cuts AI agent token costs by up to 65% on long horizon engineering tasks —and 3.5 Pro is on the way

Published by skeptic

Google DeepMind today released three new proprietary AI models it says are among its most token-efficient yet: Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber. The models aim to make AI agents faster, smarter, and cheaper at scale. Google is pricing Gemini 3.6 Flash at $1.50 per one million input tokens and $7.50 per one million output tokens through its application programming interface (API), while Gemini 3.5 Flash-Lite costs a staggeringly cheap $0.30/$2.50 per million tokens in/out. Compare that to the $1.50/$9.00 per 1M tokens for Gemini 3.5 Flash, and the $2/$12 for Gemini 3.1 Pro Preview, and the savings are considerable. However, Google’s prior generation Gemini 3.1 Flash-Lite still remains the search giant’s “most cost-efficient” model at $0.25/$1.50 per 1M tokens. Yet, it remains 2X slower than the new, more expensive Gemini 3.5 Flash-Lite, giving those enterprises who value speed more “bang” for their buck. VB Frontier AI Model API Pricing Comparison Chart (Late July 2026 Shortlist) Model Input ($/1M) Output ($/1M) Total ($/1M) Source MiMo-V2.5 Flash $0.10 $0.30 $0.40 Xiaomi …