All news

AAlex Heath12:53

AI score38

Alex Heath says Granola's free voice input is very useful

Granola has announced a new voice input feature: unlimited words, free, and usable in any app. Alex Heath says he often uses Granola to handle meetings, and now no longer needs to install a separate dictation app; he has used it for voice input on his Mac all day, and the experience has been smooth.

Quoted postGranola · @meetgranola

Granola now does voice typing. Unlimited words, for free. Granola already writes your meeting notes. Now it can do the typing for you, in any app. https://t.co/et60VtwWA0

#Product updates#Speech

SStepFun12:53First-party

AI score15

StepFun and Together Host Step 5 Preview Launch Night

StepFun and Together Compute jointly hosted the Step 5 Preview Launch Night, where the event discussed long-horizon agents, tool-calling reliability, inference costs, and what AI breakthroughs mean for the next generation of startups. The panelists included Dan Fu (Together Compute), Mar Hershenson (Pear VC), Saoud Rizwan (Cline), and Joey Wang (Harvey).

#Industry news#Agents#English text is being prepared.

TTencent Hunyuan12:53First-partyResearch

AI score62

Tencent Hunyuan Releases Hy-Nomas Technical Report and Open-Sources Hy-Nomas-Bench

Tencent Hunyuan has released the technical report for Hy-Nomas, an interactive video world model, and has open-sourced Hy-Nomas-Bench for reproducible research. Officially, Hy-Nomas was also used to reproduce Genie 4's examples and a comparison video was released, along with an online demo that requires no waitlist. Earlier introductions stated that Hy-Nomas supports free interaction with text prompts and camera trajectories, long-horizon generation with world memory, lightweight LoRA adaptation for fields such as robotics, and optimization for inference on consumer-grade GPUs.

Quoted postTencent Hy · @TencentHunyuan

🚀 Introducing Hy-Nomas: An Open Source Free-Form Interactive Video World Model Key Features: * Free-form interaction with text prompts and camera trajectories * Long-horizon generation with world memory * Lightweight LoRA adaptation for specialized domains such as robotics * Optimized for Inference on Consumer GPUs Test it and drop feedback anytime！ 🔗 Website: https://t.co/5qYjoFhtQU 💻 GitHub:https://t.co/uW5RiXbubG 🤗 Hugging Face: https://t.co/okOLzAUGjR 🎥 Demo: https://t.co/0XgSfYccLa #Hy#WorldModels #VideoGeneration #GenerativeAI

#Model releases#Video#Open-source ecosystem

KKarl's AI Watts12:53

AI score4

Karl's AI Watts: Phone and computer have both become touch screens

The author posted photos saying that their phone and computer are now touch screens too, and jokingly asked whether Apple is satisfied. The main post did not specify the exact devices or method of operation.

#English text is being prepared.

YYuchen Jin12:53

AI score8

Yuchen Jin marvels at the hand-written image inference engine from a year ago

Yuchen Jin shares a conversation with a friend: a year ago, they built an ultra-fast inference engine for image models, and the entire engine was written by hand. He exclaimed in response, "Holy shit. Just ONE year ago."

#English text is being prepared.#Reasoning

YYuchen Jin12:53Opinion

AI score65

Yuchen Jin: The OpenAI vs. Anthropic subscription battle is about tokens per dollar

Yuchen Jin commented that the focus of the OpenAI vs. Anthropic subscription competition is no longer on whose model is best, but on who offers more tokens per dollar. The Tibo post he cited announced that usage limits for all Pro 100/200/500 tiers and Business Premium are doubled, the usage limit for Pro 200 will no longer be reduced, and Work and Codex usage remain unchanged.

Quoted postTibo · @thsottiaux

Day 25/ We are doubling all usage limits for all Pro 100/200/500 tiers and Business Premium, effective today. We heard your kind feedback and won’t be reducing usage limits for Pro 200, so your Work and Codex usage will stay the same. We’re also going to continue offering extended limits for dots for now, and usage of your dot will still continue to be separate from your default usage. And we'll of course also keep shipping better models and other efficiency improvements that increase your effective usage over time.

Same story · 9 reports

#English text is being prepared.#OpenAI#Anthropic

YYuchen Jin12:53

AI score10

Silicon Valley rumor: safety researchers' salaries will surpass AI researchers'

Yuchen Jin relays a claim circulating in Silicon Valley: the income of safety researchers will soon exceed that of AI researchers.

#English text is being prepared.#Safety / alignment

MManus12:53First-party

AI score3

Manus tweets: Start building your business right now

The official Manus account posted the tweet "Build your business today," along with a link, without providing further product details.

#English text is being prepared.

KKrea AI12:53First-party

AI score23

Krea 3 to Welcome Its First Expert Model Tomorrow

Krea announced that Krea 3 will welcome its first expert model tomorrow, and said that a final clue remains in the video.

#Product updates

KKrea AI12:53First-party

AI score8

Krea and Henning Larsen Discuss AI in Architecture in New York

Krea and Henning Larsen co-hosted an event this month in New York, attended by more than 100 architects and designers, exploring the significance of AI for architecture through installations and dialogue.

#Industry news

KKrea AI12:53First-partyProduct release

AI score51

Krea Launches Ad Studio, Powered by the Krea 3 Expert Model to Generate Ads

Krea has released Ad Studio, powered by the new Krea 3 expert model designed for advertising. Users can paste a brand link to generate dozens of ads while maintaining consistency with the brand and product; it is currently available for free trial.

#Product updates#Image generation#OpenAI

KKrea AI12:53First-partyProduct release

AI score55

Krea Launches Ad Studio, Generating Ads Powered by the Krea 3 Expert Model

Krea has released Ad Studio, powered by the new Krea 3 expert model built for advertising. Users can paste a brand link to generate dozens of ads while maintaining consistency with the brand and product; it is currently available for free trial.

#Product updates#Image generation#OpenAI

GGrok12:53First-party

AI score45

Grok API Adds guidelines Content Moderation Parameter

Grok officially announced that text API requests can now include a `guidelines` parameter to further control how the Grok API handles specific content categories. When generating responses, Grok will follow these guidelines using its own wording, and this feature is available to all API customers.

Same story · 2 reports

#Product updates#English text is being prepared.#OpenAI

GGeneralist12:53First-partyProduct release

AI score50

Generalist Expands Access to GEN-1.5 and Opens Waitlist

Generalist announced that it is expanding access to GEN-1.5, its foundation model for the physical world, and has opened a waitlist for users to sign up and queue. The company also provided a detailed introduction link for GEN-1.5, as well as demonstrations of other tasks the model can complete.

Same story · 2 reports

#Product updates#Model releases#Embodied AI

GGeneralist12:53First-partyProduct release

AI score53

Generalist Expands Access to GEN-1.5 and Opens Waitlist Applications

Generalist announced that it is expanding access to GEN-1.5, its foundation model for the physical world, and users can register via a link to join the waitlist. The tweet also provides links to demonstrations of other tasks by the model, as well as a detailed page introducing GEN-1.5.

Same story · 2 reports

#Model releases#Embodied AI#OpenAI

DDemis Hassabis12:53Model releases

AI score72

Demis Hassabis announces the release of Gemini 4 Carbon, saying it achieves SOTA on multiple benchmarks

Demis Hassabis announces the release of Gemini 4 Carbon, saying it achieves SOTA on several key benchmarks and combines frontier capabilities with in-depth safety and security research. The tweet does not give specific benchmark names, scores, or release method.

Same story · 34 reports

#Model releases#Evaluations / benchmarks#Safety / alignment

DDemis Hassabis12:53

AI score0

Demis Hassabis posted, saying No.1 💎

Demis Hassabis posted on X a post containing only "No.1 💎" and an image link, without stating what or whom it referred to or the context.

#English text is being prepared.

CCognition12:53First-party

AI score60

Cognition Announces Acquisition of Corridor

Cognition announced the acquisition of Corridor and called securing global software security the most important issue in the technology field today. The tweet also includes links to announcements posted by jackhcable and AshwinRamaswami.

Same story · 3 reports

#Industry news#Safety / alignment

CCognition12:53First-party

AI score53

Cognition announces acquisition of corridor

Cognition announces the acquisition of corridor, stating that ensuring global software security is the most important issue in the technology field today. Cognition also shared links to the acquisition announcement posted by jackhcable and AshwinRamaswami.

Same story · 3 reports

#Industry news#Safety / alignment#Agents

CCognition12:53First-party

AI score47

Cognition Announces FrontierCode 2.0 Results for Both Gemini 4 Versions

Cognition evaluated Gemini 4 Carbon and Gemini 4 Argon on its in-house coding benchmark FrontierCode 2.0. Gemini 4 Carbon scored 40%, close to SOTA, at a cost of $4.04 per task, trailing Claude Opus 5.5, GPT-6 Astra, and Claude Sonnet 5.5; Gemini 4 Argon scored 26.3%, at a cost of $6.41 per task, ahead of Muse Wonder 1.0 but behind Claude Haiku 5.5 and Grok 4.7.

#Evaluations / benchmarks#English text is being prepared.#Google