AI models

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

Alex Dudov Alex Dudov AI models
2 days ago

Google DeepMind Launches Agentic Video Understanding for Gemini

Google DeepMind has introduced Agentic Video Understanding for Gemini, replacing fixed video ingestion with an adaptive, tool-driven retrieval pipeline.

Image source: aigazine.net
Usman Salis Usman Salis AI models
5 months ago

OpenAI's $122 Billion Superapp Strategy Signals a Major AI Platform Shift

OpenAI is building a unified AI superapp combining ChatGPT, coding, and agent tools. The move reflects a broader strategy to drive adoption through a single integrated platform.

Eseandre Mordi Eseandre Mordi AI models
5 months ago

Grok 4.20 Multi-Agent Beta Ranks #7 in Search Arena Leaderboard

xAI's Grok 4.20 Multi-Agent Beta has entered the Arena leaderboard with a top-10 Search Arena position, while claiming #3 in Medicine and Healthcare among leading AI systems.

Peter Smith Peter Smith AI models
5 months ago

Claude Code Leak Reveals 4 Key Features Powering Autonomous AI Systems

A reported Claude Code leak reveals how autonomous AI workflows operate beyond simple prompting. The insights highlight growing adoption of agent-based systems that act, decide, and transact without human oversight.

Peter Smith Peter Smith AI models
5 months ago

OpenAI Codex Adds Claude Code CLI Integration in Cross-Platform Move

OpenAI has launched a plugin enabling Codex models to run inside Anthropic's Claude Code CLI. The update reflects growing interoperability across AI developer tools and signals a broader shift toward flexible, cross-platform workflows.

Saad Ullah Saad Ullah AI models
5 months ago

AI Coding Boom: Claude Code NPM Downloads Surge 70x, Codex Installs Hit 5M+

New AI coding tools are seeing explosive adoption, with usage accelerating sharply after major model launches across both platforms.

Usman Salis Usman Salis AI models
5 months ago

Qwen3.5-Omni Challenges Gemini Across 100M+ Hours of AI Training Data

Qwen3.5-Omni enters the omni-modal AI arena with strong benchmark results across audio, video, and speech tasks - putting direct pressure on Google's Gemini lineup.

Peter Smith Peter Smith AI models
5 months ago

Qwen3.5-Omni: Alibaba's New Multimodal AI Handles Text, Audio, Video and 74 Languages

Qwen3.5-Omni: Alibaba's New Multimodal AI Handles Text, Audio, Video and 74 Languages

Peter Smith Peter Smith AI models
5 months ago

META's "Avocado" AI Models: 3 Variants Now in Internal Testing

Meta is reportedly testing a new family of AI models under the codename "Avocado," including multiple variants focused on efficiency, reasoning, and conversational tasks.

Usman Salis Usman Salis AI models
5 months ago

Ollama Launches Pi Coding Agent for AI Workflows

Ollama has introduced Pi, a minimal coding agent designed for customizable AI workflows. The tool expands development capabilities within the Ollama ecosystem.

Usman Salis Usman Salis AI models
5 months ago

OPUS Improves LLM Training Efficiency by 8x

Researchers introduced OPUS, a dynamic data selection method for LLM pre-training that delivers up to 8x computation reduction while boosting benchmark accuracy by 2.2% on average.

Usman Salis Usman Salis AI models
5 months ago

Microsoft's VibeVoice: Open-Source Voice AI That Handles 90-Minute Multi-Speaker Audio

Microsoft's VibeVoice combines long-form speech recognition and synthesis in a single open-source framework - processing up to 60-minute conversations and generating up to 90 minutes of consistent multi-speaker audio.

Eseandre Mordi Eseandre Mordi AI models
5 months ago

OpenAI Codex Gallery Launches Ready-to-Run Dev Workflows for Instant AI Coding

OpenAI has released a Codex Use Cases gallery, giving developers access to ready-made workflows and prompts directly inside its coding environment.

Usman Salis Usman Salis AI models
5 months ago

GPT-5.4 Hits 95% on 2026 USAMO Benchmark - a Sharp Leap from Prior Year

GPT-5.4 achieved 95% accuracy on the 2026 USA Math Olympiad benchmark, marking a sharp improvement from prior-year results.

Victoria Bazir Victoria Bazir AI models
5 months ago

Claude AI Breakthrough Raises New LLM Security Risks in 2025

A new study shows AI agents can autonomously create advanced jailbreak attacks, outperforming over 30 human-designed methods - and hitting up to 100% success rates against specific models.

Usman Salis Usman Salis AI models
5 months ago

LangChain Releases AI Agent Evaluation Checklist for Faster, More Reliable Deployments

LangChain's new structured framework walks teams through every step of building, testing, and shipping AI agents - from reading traces to handling production failures.

Eseandre Mordi Eseandre Mordi AI models
5 months ago

Anthropic Unveils Capybara AI: A New Model Tier That Goes Beyond Claude Opus

Anthropic introduces a new AI model tier called Capybara, surpassing its Opus models. The system shows major performance gains across coding, reasoning, and cybersecurity.

Eseandre Mordi Eseandre Mordi AI models
5 months ago

Gemini 3.1 Flash Hits 95.9% in Speech AI Benchmark

Google's Gemini 3.1 Flash Live Preview ranks second in a key speech reasoning benchmark. The model introduces configurable thinking levels balancing performance and latency.

Eseandre Mordi Eseandre Mordi AI models
5 months ago

GOOGL: Gemini 3.1 Flash Live Hits 90.8% on ComplexFuncBench With 70-Language Support

Google unveils Gemini 3.1 Flash Live, a real-time voice AI model with strong benchmark performance and expanded capabilities for agent-based applications.

Saad Ullah Saad Ullah AI models
5 months ago

AI Coding Tool Cuts Claude Token Use by 49x

A new open-source tool significantly reduces Claude Code token usage by restructuring how large codebases are analyzed.

Posts feed
WeatherNext 3: Google’s New AI Weather Model Delivers 5× Sharper Forecasts and Hourly Global Updates
21 hours ago
Qualcomm Adreno Neural Fusion: What It Changes for Mobile Game Developers
1 day ago
Google DeepMind Launches Agentic Video Understanding for Gemini
2 days ago
Gemini 3.8 Flash Reportedly Set to Launch Today
3 days ago
Linux Moves Toward Native Thunderbolt Support on Apple M1, M2 and M3 Macs
4 days ago
Anthropic Opens MHS Preview to Connect AI Agents With Physical Hardware
1 week ago
OpenAI's $122 Billion Superapp Strategy Signals a Major AI Platform Shift
5 months ago
BitVLA Shrinks Robot AI by 11x While Boosting Speed 4.4x
5 months ago
LFM2.5-350M Debuts With Strong Benchmark Results Across 6 Evaluation Suites
5 months ago
Grok 4.20 Multi-Agent Beta Ranks #7 in Search Arena Leaderboard
5 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