Model Performance

Twitter

Facebook

Copy Link

Key Facts

Key PlatformsLMSYS Arena, Open LLM Leaderboard

Key BenchmarksMMLU, HumanEval, GSM8K, ARC

Top News

Cellarity Unveils AI-Powered ToxPredictor Model to Revolutionize Drug Safety Testing

Cellarity has published groundbreaking research in Nature Communications introducing ToxPredictor, an AI model that predicts drug-induced liver injury with 88% sensitivity. The company has made the model and DILImap dataset publicly available to advance drug safety evaluation.

Science2 sources

Cellarity Unveils AI-Powered ToxPredictor Model to Revolutionize Drug Safety Testing

Science2 sources

OpenAI Releases GPT-5.1 with Customizable Personalities Amid Growing Legal Pressures

OpenAI introduces GPT-5.1 Instant and Thinking models featuring eight personality presets and improved performance, while facing mounting copyright lawsuits and regulatory scrutiny that could impact future operations.

Technology34 sources

OpenAI Releases GPT-5.1 with Customizable Personalities Amid Growing Legal Pressures

Technology34 sources

NVIDIA Blackwell Ultra Dominates MLPerf Training v5.1, Sets 10-Minute Record for Llama 3.1 405B

NVIDIA's Blackwell Ultra architecture sweeps all seven MLPerf Training v5.1 benchmarks, delivering unprecedented AI training performance with its GB300 NVL72 system. The platform achieves a record 10-minute training time for Llama 3.1 405B using breakthrough NVFP4 precision technology.

Technology2 sources

NVIDIA Blackwell Ultra Dominates MLPerf Training v5.1, Sets 10-Minute Record for Llama 3.1 405B

Technology2 sources

Quantum Computing Shows Promise for Kidney Disease Detection Despite Current Limitations

Florida Atlantic University researchers compared classical and quantum machine learning approaches for detecting chronic kidney disease, finding that while classical methods currently outperform quantum algorithms, quantum computing holds significant potential for future healthcare diagnostics.

Health2 sources

Quantum Computing Shows Promise for Kidney Disease Detection Despite Current Limitations

Health2 sources

NHS Trials AI Tool to Cut Prostate Cancer Diagnosis Time from Weeks to Hours

The NHS is launching a major trial of AI technology across 15 hospitals to analyze MRI scans for prostate cancer, potentially reducing diagnosis times from weeks to a single day while addressing radiologist shortages and improving patient outcomes.

Health2 sources

NHS Trials AI Tool to Cut Prostate Cancer Diagnosis Time from Weeks to Hours

Health2 sources

Meta Releases Omnilingual ASR: Open-Source Speech Recognition for 1,600+ Languages

Meta's Superintelligence Labs debuts an open-source automatic speech recognition system supporting over 1,600 languages, including 500 previously untranscribed low-resource languages. The system uses zero-shot learning to extend coverage to potentially 5,400+ languages.

Technology2 sources

Meta Releases Omnilingual ASR: Open-Source Speech Recognition for 1,600+ Languages

Technology2 sources

AI Model Revolutionizes Drug Safety by Predicting Human Toxicity Before Clinical Trials

Researchers at POSTECH have developed a machine learning model that predicts drug toxicity in humans by analyzing biological differences between animal models and humans, potentially preventing dangerous clinical trial failures.

Health2 sources

AI Model Revolutionizes Drug Safety by Predicting Human Toxicity Before Clinical Trials

Health2 sources

SoundHound AI Claims 40% Performance Edge Over Big Tech Giants as Revenue Surges 68%

SoundHound AI reports record Q3 revenue growth and raises 2025 outlook while CEO claims the company outperforms industry giants by 35-40% in accuracy. The voice AI company is expanding into new verticals including automotive, healthcare, and voice commerce.

Business3 sources

SoundHound AI Claims 40% Performance Edge Over Big Tech Giants as Revenue Surges 68%

Business3 sources

Google's Nano Banana 2 and Gemini 3 Pro Set to Revolutionize AI Image Generation and Coding

Google is preparing to launch two major AI models: Gemini 3 Pro optimized for coding with a 1 million context limit, and Nano Banana 2 featuring multi-stage image generation with native 2K support and 4K upscaling capabilities.

Technology8 sources

Google's Nano Banana 2 and Gemini 3 Pro Set to Revolutionize AI Image Generation and Coding

Technology8 sources

Chinese AI Startup Moonshot Releases Open-Source Model That Claims to Outperform GPT-5

Beijing-based Moonshot AI has released Kimi K2 Thinking, an open-source reasoning model that reportedly outperforms OpenAI's GPT-5 and Anthropic's Claude Sonnet 4.5 on key benchmarks, costing only $4.6 million to train compared to billions spent by US competitors.

Technology7 sources

Chinese AI Startup Moonshot Releases Open-Source Model That Claims to Outperform GPT-5

Technology7 sources

YouTube's AI Age Verification System Launches Second Wave, Sparking User Privacy Concerns

YouTube's AI-powered age verification system has rolled out a second wave of restrictions, incorrectly flagging adult users as minors and forcing them to choose between content limitations or sharing personal identification data.

Technology4 sources

YouTube's AI Age Verification System Launches Second Wave, Sparking User Privacy Concerns

Technology4 sources

AI Benchmarks Under Fire: Oxford Study Reveals Widespread Scientific Flaws in Model Testing

A comprehensive Oxford study exposes critical flaws in AI benchmarking methods, finding that 84% of tests lack scientific rigor and many fail to accurately measure claimed capabilities like reasoning and safety.

Science5 sources

AI Benchmarks Under Fire: Oxford Study Reveals Widespread Scientific Flaws in Model Testing

Science5 sources

Microsoft's AI Agent Marketplace Study Reveals Critical Flaws in Autonomous Shopping Systems

Microsoft researchers tested AI agents in a simulated marketplace and found they struggle with basic tasks, are easily manipulated, and perform poorly when given too many options, raising serious questions about their readiness for real-world deployment.

Science6 sources

Microsoft's AI Agent Marketplace Study Reveals Critical Flaws in Autonomous Shopping Systems

Science6 sources

Microsoft Launches MAI-Image-1: First In-House AI Image Generator Debuts in Top 10 Rankings

Microsoft has officially launched MAI-Image-1, its first fully in-house AI image generator, now available in Copilot and Bing Image Creator. The model has already achieved top 10 rankings on LMArena and is being praised for its photorealistic quality and improved hand generation.

Technology3 sources

Microsoft Launches MAI-Image-1: First In-House AI Image Generator Debuts in Top 10 Rankings

Technology3 sources

AI Shows Promise But Falls Short of Human Accuracy in Lie Detection Study

A comprehensive Michigan State University study reveals that while AI can sometimes detect human deception, it performs inconsistently and shows significant bias, falling short of human accuracy in most scenarios.

Science4 sources

AI Shows Promise But Falls Short of Human Accuracy in Lie Detection Study

Science4 sources

Prev 1…151617 18 19 Next