OpenAI introduced GPT-6.1 Sol during its 2026 Dev Day, drawing attention for its ability to match the performance of Astra, a prominent AI system, at just one-fifth of the cost. According to Universe of AI, the model is priced at $2 per input and $10 per output, making it a more accessible option for users […] The post OpenAI’s GPT-6.1 Sol Matches Astra Capabilities at Dev Day 2026 appeared first on Geeky Gadgets. Source link
Twenty-five leading mathematicians signed an open letter arguing that AI labs are threatening their intellectual work as they seek to one-up each other with solutions to famous math problems. Each signatory has been awarded the Fields Medal, considered the most prestigious prize in mathematics. This week, NYU professor Tristan Buckmaster accused OpenAI of pressuring him not to credit a collaborator who works for Anthropic for solving an important math problem, and wondered if the company had used their work with Codex to produce its own groundbreaking proof over a marathon weekend of inference. On Thursday, OpenAI withdrew its sponsorship of a math event at CalTech after the company was criticized by researchers at the university. While the ability of AI models to solve the world’s outstanding mathematical challenges could be a boon to humanity, the signatories of the new letter argue that will only be the case if those solutions can be understood and communicated by the math community and, ultimately, the rest of the world. “Often these solutions are announced in a rush, leaving …
OpenAI plans to release an AI smart speaker in 2027, according to a Bloomberg report on Thursday. It’ll reportedly be palm-size and doughnut-shaped and have a wireless design that will let people carry it around. The report, which cited unnamed people familiar with the project, also indicates that the small speaker will cost $300 to $400, a steep price compared with around $100 to $200 for most smart speakers (except for high-end Bose and Sonos options). The device would work like ChatGPT on your phone, but it would use more advanced models that can learn about the user and talk with them in tailored ways over time. A representative from OpenAI did not immediately respond to a request for comment. The report also says the speaker will have parts that move to signal how it’s interacting, something usually left to screen icons. It will have indicator lights plus a camera system and sensors to help it collect and process information about its environment. But OpenAI faces a potential legal roadblock to making the device available to consumers. …
OpenAI has introduced GPT 5.6, a release featuring three specialized models, Soul, Terra and Luna, tailored for different use cases. Soul uses its “Soul Ultra” mode to handle complex reasoning tasks, Terra offers a versatile option for general-purpose applications and Luna is optimized for high-speed, large-scale operations. According to AI Grid, these models not only surpass competitors like Claude Mythos 5 in benchmarks such as Terminal Bench but also prompt discussions around cybersecurity risks and ethical considerations due to their advanced functionalities. Explore the performance benchmarks of GPT 5.6, including a 40% cost reduction enabled by Cerebras chip integration. Learn how OpenAI is addressing regulatory challenges by limiting access to these models and collaborating with government agencies to mitigate risks like hallucinations and unintended outputs. This feature also examines the balance OpenAI is striking between advancing AI capabilities and making sure responsible deployment. Breaking Down the GPT 5.6 Variants TL;DR Key Takeaways : OpenAI has launched GPT 5.6, featuring three specialized variants, Soul, Terra and Luna, each designed for distinct use cases like advanced reasoning, …
OpenAI has made a significant update to its most widely used language model, GPT-5.5 Instant, which is the default in the free version of ChatGPT. The company announced the upgraded version of GPT-5.5 Instant yesterday on X, calling it “much more fun to talk to” and saying it is “better at understanding the intent behind a question and adapting its response accordingly,” as well as offering improvements in shopping results, local recommendations, and handling “complex constraints.” However, it has not yet provided any benchmarks or numerical results to quantify these claims. The company said the updated GPT-5.5 Instant was rolling out first to paid ChatGPT subscribers and then to free users as of today, June 25. OpenAI also updated its chat-latest API alias, which points to the latest GPT-5.5 Instant model currently used in ChatGPT, while continuing to recommend the separate gpt-5.5 model for production API usage. That distinction matters, but it should not obscure the main news: this is primarily a ChatGPT-side update to GPT-5.5 Instant, not a new release of the broader GPT-5.5 …
OpenAI released its latest model, GPT-5.5, on April 23, just a week after Anthropic introduced Claude Opus 4.7. As the two leading models from the two leading AI labs, we wanted to see how the new models compare. Spoiler alert: We think Claude Opus 4.7 has an edge on advanced and agentic coding, but GPT-5.5 performs better on most benchmarks. SEE ALSO: Anthropic says Claude Opus 4.7 has a 92% honesty rate, less sycophancy Want to learn more about getting the best out of your tech? Sign up for Mashable’s Top Stories and Deals newsletters today. GPT-5.5 and Opus 4.7: Leaderboards GPT-5.5 isn’t yet ranked on all AI leaderboards yet, but it should be very competitive with Claude Opus 4.7. On the leaderboards of verified benchmark tests such as Arc Prize, GPT-5.5 beats Opus 4.7 (more on this below). On the popular Arena leaderboard, which is based on user testing, Claude Opus 4.7 Thinking has the top overall spot. Interestingly, Opus 4.7 is currently ranked below Opus 4.6, though that will likely change in time. …
Sign up to see the future, today Can’t-miss innovations from the bleeding edge of science and tech Despite promising to allocate hundreds of billions of dollars for the build out of enormous data centers, the AI industry has struggled to keep up with its lofty ambitions. According to recent reporting by Bloomberg and Ed Zitron, roughly half the data centers slated to open in the United States are either being delayed or canceled outright. Massive electric component shortages and soaring costs have slowed the infrastructure boom to a trickle, frustrating tech leaders. In the resulting morass, every AI player has been making big claims to try to keep the hype train going. But OpenAI in particular has turned braggadocio into an art form, and its latest boast is ringing a bit sad: in a memo obtained by Bloomberg, the company boasted that it’s planning to have 30 gigawatts worth of compute — enough to power over 22 million US households — by 2030, while its rival Anthropic is only planning for seven to eight gigawatts …
On Tuesday, OpenAI rug-pulled its AI video creation tool, Sora — but it didn’t feel like a loss. If anything, it felt like a small win for AI-free creativity. However, that didn’t last long, with TikTok owner ByteDance announcing the global rollout of its Seedance 2.0 video creation tool, making its CapCut AI video generation app the first platform to jump on board. CapCut’s blog post revealed that the new audio and video model, Dreamina Seedance 2.0, is already rolling out to the platform, with ByteDance stating that the model will allow creators to draft, edit, and sync video and audio content using simple text prompts, images, or reference videos. And while that’s great news for CapCut and AI video generation, it’s a step back for those who thought the AI-slop tide was just starting to turn. Related OpenAI pulled the plug on Sora without warning — even Disney didn’t see it coming Disney’s $1 billion deal with the AI slop merchant is dusted. Seedance 2.0 is going straight to CapCut Meet your favorite new …
It’s unclear what OpenAI’s motivations are. It’s not the first tech giant to embrace military contracts it had once vowed never to enter into, but the speed of the pivot was notable. Perhaps it’s just about money; OpenAI is spending lots on AI training and is on the hunt for more revenue (from sources including ads). Or perhaps Altman truly believes the ideological framing he often invokes: that liberal democracies (and their militaries) must have access to the most powerful AI to compete with China. The more consequential question is what happens next. OpenAI has decided it is comfortable operating right in the messy heart of combat, just as the US escalates its strikes against Iran (with AI playing a larger role in that than ever before). So where exactly could OpenAI’s tech show up in this fight? And which applications will its customers (and employees) tolerate? Targets and strikes Though its Pentagon agreement is in place, it’s unclear when OpenAI’s technology will be ready for classified environments, since it must be integrated with other …
Stuffing an AI chatbot into a consumer electronics device and turning out with a product people actually want has proven extremely difficult. We’ve come across creepy and widely-hated pendants designed to listen to everything you say, as well as flawed AI “pins” that turned out to be a flaming dumpster fire, leading to frustration and disbelief. Now Sam Altman’s OpenAI, which recruited former Apple design lead Jony Ive to lead its own hardware effort, is gearing up to release its would-be showstopper — and alas, it doesn’t sound like it’s managed to iterate beyond “clunky gadget that pretty much does stuff your phone already does.” As The Information reports, a team of over 200 employees has been working on a smart speaker that features a built-in camera, which will recognize faces and identify objects thanks to a dose of AI. It will reportedly retail for anywhere from $200 and $300 and ship no earlier than the beginning of next year. That’s right: the best idea that the big brains at OpenAI could cook up is …
