#ai model

105 articles

Marina Lyubimova Marina Lyubimova Press Releases
19 hours ago

Anthropic Opens MHS Preview to Connect AI Agents With Physical Hardware

Anthropic has opened a research preview of the Model Hardware Standard (MHS) to selected research labs and advanced manufacturers. The specification gives AI agents a common interface for controlling microscopes, liquid handlers, robotic arms, lasers and other programmable equipment.

Image source: aigazine.net
Eseandre Mordi Eseandre Mordi Industry
5 months ago

3DThinker AI Model Unlocks 3D Reasoning From 2D Images Without Labeled Data

3DThinker enables AI 3D spatial reasoning from 2D images without labeled 3D data, outperforming baselines across benchmarks.

Victoria Bazir Victoria Bazir Industry
5 months ago

Unsloth Studio Cuts VRAM by 70% and Trains AI Models 2x Faster With Triton Kernels

Unsloth Studio uses Triton GPU kernels for 2x faster AI training and up to 70% less VRAM.

Peter Smith Peter Smith Industry
5 months ago

AI Research Warns of "Artificial Hivemind" Across 70+ Language Models

Study of 70+ AI models finds they give nearly identical answers to open-ended prompts, raising concerns about systemic bias.

Victoria Bazir Victoria Bazir Industry
5 months ago

Alibaba Launches REG4Rec AI Model With 16.59% Accuracy Boost

Alibaba and Wuhan University unveil REG4Rec, a reasoning-enhanced AI model boosting e-commerce recommendation accuracy by 16.59% and already live on Alibaba's ad platform.

Victoria Bazir Victoria Bazir AI models
5 months ago

Aheadform F1 AI Humanoid Robot: 4 Roles, 1 Machine Built for Human Connection

Aheadform has introduced the F1 half-humanoid robot designed for companionship and social interaction. The system can dynamically switch between teacher, learning assistant, companion, and therapist roles depending on context.

Victoria Bazir Victoria Bazir AI models
5 months ago

BridgeV2W: AI Model With Embodied Masks Boosts Robot Motion Prediction

Researchers from the Chinese Academy of Sciences, UCAS, and FiveAges unveiled BridgeV2W, an AI model that enables robots to visualize future motion by generating embodiment masks. The system reportedly outperforms current state-of-the-art frameworks on robotics datasets.

Eseandre Mordi Eseandre Mordi Industry
5 months ago

DeepSeek Blocks Nvidia & AMD From New AI Model With 75 Million Downloads

DeepSeek has reportedly denied Nvidia and AMD access to its latest AI model, diverging from typical industry practice. The situation also includes unverified claims the model was trained on banned Nvidia chips.

Saad Ullah Saad Ullah Benchmarks
6 months ago

New Benchmark Tests AI Models with 55 Nonsensical Questions

A new "Bullshit Benchmark" tests how AI models handle illogical questions, revealing that many still generate confident answers to meaningless prompts instead of rejecting them outright.

Peter Smith Peter Smith Industry
6 months ago

Google Drops TimesFM 2.5 - New AI Model Hits 128-Step Forecast Horizon

Google just launched TimesFM 2.5 on Hugging Face, a specialized AI model designed for time series forecasting. The decoder-only architecture delivers solid zero-shot performance right out of the box, no fine-tuning needed.

Saad Ullah Saad Ullah AI models
6 months ago

Chinese AI Models Account for 61% of OpenRouter Token Usage

Chinese AI models dominate OpenRouter's platform, processing 61% of all tokens according to new usage data. MiniMax M2.5 leads with 2.45 trillion weekly tokens as deployment scales accelerate.

Saad Ullah Saad Ullah AI models
6 months ago

Grok Climbs to #3 in Global GenAI Rankings Just 12 Months After Launch

Grok has rocketed from an unranked early-stage launch to become the world's third most popular generative AI tool in just one year.

Usman Salis Usman Salis AI models
6 months ago

53.79% of Grok Visitors Were New Users in January 2026

Similarweb reports new users made up 53.79% of Grok visitors in January 2026, with new visitor growth rising 74.83% month over month.

Eseandre Mordi Eseandre Mordi AI models
6 months ago

MOVA AI Model Introduces Multimodal Video-Audio Generation

The open-source MOVA model generates video, speech, music, and sound together using a shared architecture designed to keep modalities synchronized.

Eseandre Mordi Eseandre Mordi AI models
6 months ago

Kimi K2.5 Hits 856 Billion Tokens, Leads OpenRouter Usage Rankings

Kimi K2.5 takes the number one spot on OpenRouter's leaderboard with real-world usage metrics showing it processed 856 billion tokens in a single week.

Victoria Bazir Victoria Bazir AI models
6 months ago

Grok-4.20 Dominates Alpha Arena Leaderboard with 34.59% Live Trading Returns

Grok-4.20 crushes the competition on Alpha Arena's live trading leaderboard, holding the top spot with real-money returns that jumped from 12% to nearly 35% in just 10 days. Four Grok variants currently occupy the top six positions.

Peter Smith Peter Smith AI models
6 months ago

AI Coding Models Show Early Signs of Recursive Self-Improvement

New commentary from AI developers suggests that modern coding models are increasingly helping to improve the tools used to build future AI systems, though full automation remains incomplete.

Victoria Bazir Victoria Bazir AI models
6 months ago

Ant Group Unveils LingBot-VLA: Universal Robot Control Model Trained on 20,000 Hours of Real-World Data

Ant Group researchers just dropped LingBot-VLA, a vision-language-action AI model that can control multiple robot types through a single system. The model crushes existing approaches in both performance and training speed across real-world robotic tasks.

Peter Smith Peter Smith AI models
6 months ago

AI Wage Models Show 40% Efficiency Gains in Early Tests

New wage-based compensation models for AI systems are showing significant efficiency improvements, with early testing revealing productivity gains of 20-50% as agents optimize their resource usage to maximize earnings.

Saad Ullah Saad Ullah Benchmarks
6 months ago

Trinity-Large Delivers 400B-Parameter Speed with 256-Expert Architecture

Fresh benchmark data reveals major performance gaps between leading language models. Trinity-Large pulls ahead with exceptional throughput and response speed thanks to its ultra-sparse mixture-of-experts design.

Posts feed
Anthropic Opens MHS Preview to Connect AI Agents With Physical Hardware
19 hours ago
OpenAI's $122 Billion Superapp Strategy Signals a Major AI Platform Shift
4 months ago
BitVLA Shrinks Robot AI by 11x While Boosting Speed 4.4x
4 months ago
LFM2.5-350M Debuts With Strong Benchmark Results Across 6 Evaluation Suites
4 months ago
Grok 4.20 Multi-Agent Beta Ranks #7 in Search Arena Leaderboard
4 months ago
Claude Code Leak Reveals 4 Key Features Powering Autonomous AI Systems
4 months ago
RISE Robot Framework Boosts Task Success Rates by 35% Using AI-Driven Imagination
4 months ago
Multi-Agent AI Systems Face 15 Emerging Social Risks, New Study Finds
4 months ago
Mistral AI Raises $830M to Build Paris AI Data Center
4 months ago
OpenAI Codex Adds Claude Code CLI Integration in Cross-Platform Move
4 months ago
Follow us on Twitter

Aigazine is an independent publication covering artificial intelligence — new models, research, the companies building them, and the industries they reshape. Published from Tbilisi, Georgia (0179, Tbilisi City, Vake District, 49 Besarion Zhghenti Street, VAT 305786600).

© 2026 Aigazine

We use cookies to improve your experience on our site and to show you relevant advertising. To find our more, read our privacy policy and cookie policy