AI models

News and updates on leading AI models - from ChatGPT and Claude to Gemini and open-source innovations.

Saad Ullah Saad Ullah AI models
9 months ago

Opus 4.5 Release Sparks Debate as AI Sector Watches Gemini 3 Pro Showdown

A leaked Epoch AI schedule hints Opus 4.5 could drop today, fueling speculation about whether it can beat Google's Gemini 3 Pro and shift the balance in the AI race.

Peter Smith Peter Smith AI models
9 months ago

Memori Launches SQL-Native Memory Engine with 80-90% Cost Reduction for LLM Agents

Memori introduces an open-source memory engine that gives AI agents long-term memory using standard SQL databases instead of expensive vector storage, cutting costs by up to 90 percent while working across all major LLM platforms.

Sergey Diakov Sergey Diakov AI models
9 months ago

AI Models Hit 98th Percentile as New Offline Test Shows Mensa-Level Scores for First Time

A new offline IQ test reveals an AI system scoring at Mensa level for the first time, jumping from a 92 IQ score just two years ago. The development mirrors solar energy's exponential growth pattern, where actual installations outpaced forecasts by over 3x

Peter Smith Peter Smith AI models
9 months ago

Gemini 3 Launch Drives Market Share Up 6.77 Points to 30%

Google released Gemini 3 and within days the platform's share of the AI-assistant market leapt from 23 % to just over 30 %. Traffic plus user engagement rose at the same time.

Saad Ullah Saad Ullah AI models
9 months ago

Gemini 3 Pro Scores 142 on Mensa Norway IQ Test

Google's Gemini 3 Pro has produced IQ scores of 130 in an offline test and 142 on the Mensa Norway benchmark placing it at the top of current AI models on human-style reasoning exams.

Usman Salis Usman Salis AI models
9 months ago

SharpaWave's 22-DOF Robotic Hand Achieves Human-Level Dexterity

Sharpa Robotics showcased its SharpaWave robotic hand performing high-precision manipulation tasks with human-like control. Recent demonstrations highlight significant progress in tactile sensing, teleoperation capabilities, and fine-motor actuation.

Eseandre Mordi Eseandre Mordi AI models
9 months ago

xAI's Grok Models Dominate OpenRouter: 300B+ Tokens/Day Leave Competitors Behind

Grok models have locked down the top three spots on OpenRouter's leaderboard, crushing past 300 billion tokens daily. Meanwhile, every competing system is stuck below 50 billion tokens—and it's not even close.

Usman Salis Usman Salis AI models
9 months ago

Codex Max Boosts GPT-5.1 Pro Code Planning Efficiency

Codex Max is improving how GPT-5.1 Pro handles complex code-planning tasks, offering a stronger workflow for large-scale software development and highlighting growing reliance on advanced AI systems in engineering environments.

Saad Ullah Saad Ullah AI models
9 months ago

OpenAI Energy Demand Surges: AI Giant May Match UK and Germany in 5 Years

New projections suggest OpenAI's electricity usage could surpass UK and Germany consumption within five years and potentially exceed India's within eight years, sparking concerns about AI sustainability and energy policy.

Peter Smith Peter Smith AI models
9 months ago

GLM Targets 30B-Parameter AI Model for 2025 as Chinese Players Push for Market Share

GLM is gearing up to launch a new 30-billion-parameter AI model in 2025, moving beyond its massive 355-billion-parameter system as Chinese AI companies race to expand their portfolios and grab bigger slices of the market.

Peter Smith Peter Smith AI models
9 months ago

$1 Billion Figure AI Sued Over Robots That Could Fracture Human Skulls

Figure AI is facing a whistleblower lawsuit from its former Head of Product Safety, who claims the company rushed unsafe humanoid robots to market after raising over $1 billion. The complaint alleges robots generated enough force to fracture skulls and that the company retaliated when safety concerns were raised.

Eseandre Mordi Eseandre Mordi AI models
9 months ago

Grok 4.1 Fast Hits No. 1 on OpenRouter with 275 Billion Tokens, Overtakes Google's Gemini

xAI's Grok 4.1 Fast has grabbed the top spot on OpenRouter's Trending Leaderboard, pushing Google's Gemini down to second place. The numbers show a massive spike in usage across the platform.

Saad Ullah Saad Ullah AI models
9 months ago

Study Reveals LLMs Contradict Their Own Beliefs by Up to 30%

Recent research shows large language models frequently update their beliefs incorrectly and make decisions that directly contradict their stated probabilities, raising concerns about their reliability in real-world applications.

Sergey Diakov Sergey Diakov AI models
9 months ago

xAI's Grok Models Claim No.1 and No.2 Spots on Programming Leaderboard, Beating Google and Anthropic

xAI's Grok Code Fast 1 and Grok 4.1 Fast have topped OpenRouter's programming leaderboard, pushing ahead of Google's Gemini and Anthropic's Claude models with strong performance in Python and JavaScript tasks.

Artem Voloskovets Artem Voloskovets AI models
9 months ago

Gemini 3 Pro Shows Lower Error Rates but Still Trails Claude Sonnet 4.5 at 1.5%

Fresh PostHog data reveals Gemini 3 Pro significantly cutting tool error rates compared to Gemini 2.5 Pro, though Claude Sonnet 4.5 remains the most reliable model with just 1.5% errors.

Saad Ullah Saad Ullah AI models
9 months ago

Codex CLI Stands Out as Only Agent Beating Default Terminus 2 Benchmark

New Terminal Bench results show Codex CLI as the only evaluated agent outperforming the baseline Terminus 2 model, revealing a significant performance gap among leading terminal-focused AI systems.

Peter Smith Peter Smith AI models
9 months ago

Gemini 3 Delivers 3,000% Higher Profit Generation Than Competing AI Models

Google's Gemini 3 is crushing the competition with massive performance gains in real-world economic testing. The model's profitability advantage signals a major shift in how AI will reshape automation and business productivity.

Peter Smith Peter Smith AI models
9 months ago

Moonshot's Kimi K2 Models Hit 60.2 on BrowseComp as K2 Thinking Turbo Sets New Agentic Benchmark

Moonshot AI's Kimi K2 Thinking and K2 Thinking Turbo models are outperforming open-weights competitors on reasoning and multi-step tasks, with the systems using smart planning cycles and tool integration while staying efficient.

Saad Ullah Saad Ullah AI models
9 months ago

AI Models Learn to Deceive: Anthropic Study Reveals 50% of Tests Show Hidden Deception

New Anthropic research reveals AI models trained on coding exploits developed widespread deceptive behaviors, with 50% of tests showing intentional concealment. A single prompt adjustment eliminated the dangerous pattern.

Eseandre Mordi Eseandre Mordi AI models
9 months ago

Sam Altman Says Next AI Leap Targets Multi-Month Tasks With GPT-5

Sam Altman revealed a major shift in AI development, with models advancing from seconds-long tasks to 5-hour workflows in GPT-5. The next goal is enabling AI systems to operate across months or even years using integrated enterprise context.

Posts feed
Anthropic Opens MHS Preview to Connect AI Agents With Physical Hardware
14 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