A decade of AI stack innovation now culminates in record-breaking benchmarks for model speed and latency

WASHINGTON, Sept. 25, 2025 /PRNewswire/ — Clarifai, a global leader in AI and pioneer of the full-stack AI platform, today announced the launch of Clarifai Reasoning Engine, a breakthrough in inference performance designed specifically for agentic use cases and reasoning models.

End-to-End Response Time. Lower is better.

Since 2013, Clarifai has pioneered the modern AI stack from computer vision customization to end-to-end model operations platforms. Leveraging years of experience operating frontier models in production, Clarifai’s team developed the new Clarifai Reasoning Engine with optimized kernels and novel techniques that dynamically adapt to workloads, improving generation speed over time without compromising accuracy.

The launch of Clarifai Reasoning Engine follows a recent benchmarking by Artificial Analysis against top API providers for the performance of OpenAI’s gpt-oss 120B model. The Clarifai Reasoning Engine recorded new industry-best records for throughput and latency on GPUs (and even bested some specialized ASIC chips from other providers).

After recently beginning evaluation by Artificial Analysis of Clarifai’s hosted gpt-oss-120b model, it now achieves new records of speed, serving over 500 tokens per second with a time to first token of 0.3 seconds. In a subsequent round of tests, the Clarifai Reasoning Engine outperformed not only all GPU-based inference implementations, but also specialized non-GPU accelerators, proving for the first time that GPU performance can match—and in some cases surpass—non-GPU architectures.

These results mark a milestone in AI operations. Combined with Clarifai’s industry-leading compute orchestration technology, Clarifai Reasoning Engine delivers the speed, flexibility, efficiency, and reliability that enterprises and developers need to scale intelligent applications without being locked into a single hardware vendor.

“Agentic AI and reasoning workloads burn through tokens rapidly. They require high throughput, low latency and low prices to drive viable customer use-cases,” said Matthew Zeiler, Founder & Chief Executive Officer at Clarifai. “With the Clarifai Reasoning Engine, developers can unlock a new era of speed and responsiveness with inference more affordable than the best industry offerings today. We’re also using our technology and AI expertise to help model builders deliver breakthrough inference performance for their custom models on standard GPUs that is now competitive with specialty non-GPU architectures for the first time ever in our industry, with agent-friendly pricing.”

Clarifai’s latest release is optimized for agentic AI, purpose-built to accelerate the latest reasoning models and automation tasks across industries. Note that Clarifai’s performance is not one and done, but rather with adaptive performance that learns from workload behavior, the platform continually improves speed without compromising accuracy. These improvements compound over time, especially for the repetitive tasks that are common in agentic workflows. Starting today, customers can engage with Clarifai’s AI experts to apply these optimizations to their own models, further enhancing performance and economics.

About Clarifai
Clarifai is a global leader in AI and the pioneer of the full-stack AI platform that helps organizations, teams, and developers build, deploy, and operationalize AI at the highest scale and availability. Clarifai’s cutting-edge AI platform supports today’s modern AI technologies like Large Language Models (LLMs), Large Vision Models (LVMs), Retrieval Augmented Generation (RAG), automated data labeling, high-volume production inference, and more. Founded in 2013, Clarifai is available in cloud, on-premises, or hybrid environments and has been used to build more than 1.5 million AI models, serving more than 500,000 users across 170 countries. Learn more at www.clarifai.com.

For more information or media requests, contact: pr@clarifai.com.

Latency vs. Output Speed. Size represents price. Lower

Output Speed vs. Price.

• New Gameplay Trailer Revealed During State of Play; Pre-order Package for PlayStation® 5 Now Available
• Tokyo Game Show Press Conference Unveiled Closed Beta Schedule 

SEOUL, South Korea, Sept. 25, 2025 /PRNewswire/ — Netmarble, a leading developer and publisher of high-quality games, announced its upcoming open-world action RPG, The Seven Deadly Sins: Origin, will launch globally on PlayStation® 5, Steam (PC), and mobile on January 28, 2026. The announcement was made during Sony Interactive Entertainment’s State of Play, accompanied by the debut of a new gameplay trailer. 

The newly revealed walkthrough video highlighted the game’s seamless blend of single-player storytelling and multiplayer co-op features. Players will explore the continent of Britannia while engaging in various activities such as fishing, cooking, resource gathering, and more. Multiplayer features allow players to form a party of up to five to challenge dungeons and discover hidden secrets.

Alongside the reveal, a PlayStation® 5 pre-order package is now available on the PlayStation® Store, featuring in-game items such as Draw Tickets, Hero Upgrade Materials, and Weapon Upgrade Materials.

Tokyo Game Show 2025 Press Event Reveals CBT Schedule

Following its State of Play reveal, The Seven Deadly Sins: Origin also held a press conference at Tokyo Game Show 2025, offering a deeper look at the game’s development and upcoming Closed Beta Test. The event featured live appearances by the voice actors of main characters “Tristan” and “Tioreh,” alongside the game’s key developers.

Attendees were treated to a live gameplay demonstration featuring the open world of Britannia and core combat mechanics. Additionally, a special video interview with The Seven Deadly Sins creator Nakaba Suzuki was presented, offering insight into the game’s multiverse narrative and its connection to the original series.

One of the key announcements was the Closed Beta Test (CBT), scheduled to run for one week from October 30 to November 5 2025 on PlayStation® 5 and PC. Participants will gain early access to the opening chapters, combat systems, and selected regions of the game’s open world.

The Seven Deadly Sins: Origin is based on The Seven Deadly Sins, the popular Japanese manga and anime IP that has reached 55 million copies in global sales worldwide. The Seven Deadly Sins: Origin will feature a multiverse storyline original to the game and an expansive open world across the continent of Britannia, allowing players to collect heroes from The Seven Deadly Sins and Four Knights of the Apocalypse to customize their combat style and shape their own adventure. 

Having been featured at major global gaming events including gamescom 2025 Opening Night Live and Tokyo Game Show 2025, The Seven Deadly Sins: Origin is gaining recognition as one of the most anticipated RPGs of the upcoming year.

Fans can sign up for pre-registration on Google Play, the App Store, and the official brand website. Players can also wishlist the game on PlayStation® 5 and Steam. To stay updated, follow The Seven Deadly Sins: Origin on its official YouTube, X, and Discord channels. 

©Nakaba Suzuki, KODANSHA/The Seven Deadly Sins: Revival of the Commandments Project,MBS. All Rights Reserved.
©Nakaba Suzuki, KODANSHA/The Seven Deadly Sins: Dragon’s Judgement Production Committee, TX
©Nakaba Suzuki, KODANSHA/The Seven Deadly Sins: Four Knights of the Apocalypse Production Committee.
©2025, “PlayStation”, “PS5” are registered trademarks or trademarks of Sony Interactive Entertainment Inc.
©Netmarble Corp. & Netmarble F&C Inc. All Rights Reserved.

About Netmarble Corporation

Founded in Korea in 2000, Netmarble Corporation is a leading global game developer and publisher. Through acclaimed franchises and strategic partnerships with top-tier IP holders, the company delivers innovative and engaging gaming experiences to audiences worldwide. As a parent company of Kabam, SpinX Games, Jam City, and a major shareholder of HYBE and NCSOFT, Netmarble’s diverse portfolio includes Solo Leveling:ARISE, Seven Knights Idle Adventure, Tower of God: New World, Lineage 2: Revolution, MARVEL Future Fight, Ni no Kuni: Cross Worlds and The Seven Deadly Sins: Grand Cross. More information can be found at http://company.netmarble.com.

SEOUL, South Korea, Sept. 25, 2025 /PRNewswire/ — DESILO Inc. today announced the introduction of its THOR framework, a breakthrough in privacy-safe AI that enables large language model (LLM) inference to run fully under homomorphic encryption. The company’s joint research paper with Professor Miran Kim’s research team at Hanyang University has been accepted for presentation at ACM CCS 2025, one of the world’s premier peer-reviewed venues for computer and communication security research, alongside IEEE S&P and USENIX Security.

The research delivers two key results. First, the team ran a widely used open-source model under homomorphic encryption. They did so without retraining, replacing its computations with homomorphic encryption so inputs and outputs remained encrypted throughout inference.

Second, they achieved near practical runtime. Roughly two sentences of input (128 tokens) are answered on a single GPU at deployment-relevant speed. On core matrix multiplication, performance improved by 5.3x for plaintext to ciphertext operations and 9.7x for ciphertext to ciphertext operations, setting a new state of the art in the team’s evaluation setting. To the authors’ knowledge, THOR is the first framework to run an entire existing LLM under homomorphic encryption without retraining while still reaching near practical speed.

“This CCS acceptance recognizes a breakthrough in running large language models fully under homomorphic encryption, marking an important academic milestone in privacy-preserving AI research,” said Seungmyung Lee, CEO of DESILO. “It also highlights our dual focus: driving forward homomorphic encryption research with partners like Cornami, and accelerating product development to bring trusted Privacy AI into real-world applications.”

This innovation forms the technical foundation for upcoming DESILO solutions such as Harvest™, designed to enable secure, privacy-safe analysis across multiple institutions.

Exclusive Early Prime Big Deals Offerings Available Now

SAN FRANCISCO, Sept. 25, 2025 /PRNewswire/ — In a crowded market where many smart cleaning devices promise convenience but often fall short, YEEDI stands apart with its latest innovation: the S20 Infinity. Engineered for large homes, complex surfaces, and demanding messes, the S20 Infinity marries cutting-edge technology with unparalleled endurance — letting users reclaim their time and peace of mind.

PowerBoost Charging: Designed to Solve Real User Frustrations

For many robot vacuum users, battery life and charging downtime remain the biggest frustrations. It’s common to find your cleaning interrupted halfway through a large home because the vacuum runs out of power. Long charging times mean you can’t simply run back-to-back cleaning sessions or trust your device to handle big jobs without supervision. Constantly having to monitor, schedule, or manually recharge the robot defeats the whole purpose of a “smart” cleaner — it should give you freedom, not more hassle.

YEEDI’s PowerBoost Charging is designed specifically to tackle these issues. This innovative system enables ultra-fast, in-use top-up charging that dramatically extends runtime without pausing the cleaning process.

• Rapid top-ups on the fly: Just a 3-minute charge replenishes 6% of battery capacity, enough to maintain high suction and mopping power continuously.
• Full recharge in under an hour: The S20 Infinity battery charges completely in less than 60 minutes — far faster than most competitors.
• Exceptional cleaning coverage: Thanks to PowerBoost, the S20 Infinity cleans over 10,000 square feet per session without losing power or performance.
• Smart power management: This technology eliminates the need to micromanage cleaning schedules or worry about battery interruptions, giving users a truly hands-off, worry-free experience.

In essence, PowerBoost Charging transforms how robot vacuums operate — turning battery life from a limiting factor into a seamless advantage.

PureCyclone Station: Bagless, Consistent, and Hassle-Free

The PureCyclone 2.0 vacuum station features a two-stage cyclonic airflow system, ensuring consistent suction strength with less than 1% loss at full capacity — a dramatic improvement over conventional systems that can lose over 60%. Its large 1.6L dustbin holds debris for over a month, eliminating the need for frequent emptying or messy bag replacements.

Advanced Cleaning Power: OZMO Roller 2.0 & BLAST Suction System

The OZMO Roller 2.0 delivers next-level mopping with a horizontally mounted roller that maximizes ground contact. It continuously cycles clean water while extracting dirty water, preventing smears even on stubborn spills. Operating at 200 RPM with 3800Pa suction, it easily handles dried stains and daily grime.

Paired with this is the BLAST (Boosted Large-Airflow Suction Technology) vacuum system. Featuring a 100W motor and optimized airflow, BLAST provides powerful yet quiet suction—ensuring deep cleaning without excessive noise.

Intelligent Navigation for Complex Homes

Equipped with TruePass Adaptive 4WD Climbing and TruEdge 3.0 technology, the S20 Infinity navigates thresholds up to 2.4cm high and cleans edge-to-edge with precision. Whether moving across tiles, carpets, or around delicate molding, it delivers accurate, collision-free cleaning tailored to your home’s layout.

Fall into Savings: Early Prime Big Deals Day Promotions

As the fall season kicks in, YEEDI invites consumers to upgrade their home cleaning with exclusive savings during the Early Prime Big Deals Day, running from September 22 to October 8. Featured discounts include:

• S20 Infinity – Now $1099.99 (Save $600)
• M14 PLUS – Featuring advanced OZMO Roller technology, now available at a special discounted price of $599.99
• M12 PRO PLUS – Ultra-powerful 11,000Pa suction, now only $379.99 (53% off)
• C12 PRO PLUS and other models – Also heavily discounted

Whether you’re managing a large household or seeking hassle-free smart cleaning, the S20 Infinity delivers unmatched performance and convenience.

Visit the official YEEDI online store to explore these limited-time offers and experience the future of effortless home cleaning.

About YEEDI

YEEDI is a cutting-edge service robot brand that embraces the concept of “Live Smart. Enjoy Life.” With constant upgrades, pioneering technology solutions are integrated into robotics for an enhanced consumer experience. Thus, consumers can enjoy a modern lifestyle with high standards of quality.

$100 Million Partnership Brings OpenAI Models Natively to 20,000+ Customers on the Databricks Data Intelligence Platform

SAN FRANCISCO, Sept. 25, 2025 /PRNewswire/ — Databricks and OpenAI today announced a multi-year partnership to make OpenAI models natively available within the Databricks Data Intelligence Platform and Databricks’ flagship AI product Agent Bricks. OpenAI models will be available to all 20,000+ Databricks customers across clouds. This collaboration unites two of the world’s most impactful technology companies, with a $100 million partnership, to deliver frontier AI for enterprises around the world. With Agent Bricks, organizations can build, evaluate and scale production-grade AI apps and agents on their governed enterprise data leveraging OpenAI latest models, including GPT-5, featuring state-of-the-art performance across coding, math, writing, and reasoning. Under this partnership, GPT-5 will be a flagship model for all Databricks customers.

Enterprise-Ready AI Apps and Agents
Businesses need AI agents that can analyze and act on their enterprise data, automate complex workflows and deliver reliable insights while meeting strict standards for accuracy, security and governance. Developing AI agents can often require copying data, juggling multiple tools, uncertain evaluation techniques or cost and quality trade-offs. Databricks customers now benefit from GPT-5 and OpenAI advanced natural language processing, reasoning and complex decision-making to build high-quality, governed AI agents using Databricks Agent Bricks. OpenAI models are readily available where enterprise data already lives, requiring no extra setup and no data movement and with guaranteed high-capacity processing.

Customers use the Data Intelligence Platform and Agent Bricks to find and treat diseases earlier, identify ways to become more energy efficient, detect financial fraud, accelerate application development, and much more.

Quotes:

• “We’re seeing overwhelming demand from enterprise customers looking to build AI apps and agents on their data, tailored to their unique business needs,” said Ali Ghodsi, Co-founder and CEO of Databricks. “This partnership makes it easier for enterprises to securely leverage their data and OpenAI models at scale with best-in-class governance and performance.”
• “Enterprise demand for frontier AI is accelerating, and with Databricks, we’re making its deployment even simpler without compromising the high bar for performance and production,” said Brad Lightcap, COO of OpenAI. “Our partnership with Databricks brings our most advanced models to where secure enterprise data already lives, making it easier for businesses to experiment, deploy, and scale AI agents with real impact.”
• “At Mastercard, we are focused on delivering AI solutions that make commerce safer, smarter, and more personalized. Agentic solutions also make our internal operations stronger by automating processes and optimizing performance across our systems,” said Greg Ulrich, Chief AI and Data Officer at Mastercard. “For any use case, AI agents come down to three things: quality, scale, and trust. This partnership between Databricks and OpenAI enables us to build on the strong foundations we’ve established with each firm and provides us the opportunity to build trusted AI agents that harness the latest OpenAI models—delivered with the speed, security and scale of the Databricks platform.”

Partnership Highlights:

• Frontier Models on Enterprise Data: Databricks customers can now bring OpenAI models directly to their enterprise data — accessible in SQL or via API — to run LLMs at scale, enrich their data and securely build high-quality custom agents.
• Dedicated High Capacity: Databricks customers benefit from access to high-capacity processing across the latest OpenAI models.
• Production-ready AI Agents with Agent Bricks: Agent Bricks builds production-ready AI agents by measuring the accuracy of models like GPT-5 and gpt-oss with task-specific evaluation and LLM judges. Agent Bricks then tunes and optimizes the model to deliver high-quality, domain-specific outputs.
• Unified Governance and Built-in Observability: Databricks Unity Catalog provides end-to-end governance and responsible AI controls, helping ensure compliance with security, access, and ethical standards — so teams can ship agents into production with confidence.
• Collaborative Innovation: Databricks and OpenAI technical teams will work closely together to continually improve and optimize OpenAI models for enterprise use cases, ensuring high performance and rapid iteration for AI workloads.

This partnership builds upon the companies’ existing work together, where OpenAI leverages Databricks for the processing of AI data to help generate insights that enhance product quality and elevate the ChatGPT experience. Databricks was also one of the first to host gpt-oss, the OpenAI open weight model.

Join us on November 11 (AMER) and November 12 (APJ / EMEA) for The Future of AI: Build Agents That Work, a conversation with Ali Ghodsi, Co-founder and CEO at Databricks, and Sam Altman, Co-founder and CEO at OpenAI, on where AI is heading and how to build agents that deliver real results. Register here.

About Databricks
Databricks is the Data and AI company. More than 20,000 organizations worldwide — including Block, Comcast, Condé Nast, Rivian, Shell and over 60% of the Fortune 500 — rely on the Databricks Data Intelligence Platform to take control of their data and put it to work with AI. Databricks is headquartered in San Francisco, with offices around the globe and was founded by the original creators of Lakehouse, Apache Spark™, Delta Lake, MLflow, and Unity Catalog. To learn more, follow Databricks on X, LinkedIn and Facebook.

NEW YORK, Sept. 25, 2025 /PRNewswire/ — The New York Stock Exchange (NYSE) provides a daily pre-market update directly from the NYSE Trading Floor. Access today’s NYSE Pre-market update for market insights before trading begins. 

Ashley Mastronardi delivers the pre-market update on September 25th

• Equities are fractionally lower following two straight days of losses across the major averages. Investors will focus on new developments in the AI trade, as well as key economic data out ahead of today’s open.
• The Labor Department released its latest weekly jobless claims data this morning amid a cooling labor market. Fed Chair Jerome Powell said this week that weakness in the labor market is outweighing concerns about sticky inflation.
• Today at the New York Stock Exchange, the President of South Korea, Lee Jae Myung, will ring the Opening Bell amid the 80^th session of the United Nations General Assembly.  

Opening Bell
H.E. Lee Jae Myung, President of South Korea, rings the Opening Bell

Closing Bell
Tetra Technologies (NYSE: TTI) celebrates ONE TETRA – The Path to 2030

Click here to download the NYSE TV App

Video – https://mma.prnasia.com/media2/2782150/NYSE_Market_Update_Sept_25.mp4