Alexey Shabanov

Google is testing a Projects feature for Gemini Enterprise that allows users to organize chats by topic, upload files for cross-chat reference, and set goals to guide AI responses—capabilities that Anthropic and OpenAI have already deployed.

“Author of "Google tests Projects feature for Gemini Enterprise" in Testing Catalog”

Anthropic is developing a web-based version of Claude Code with GitHub integration and managed sandboxes to compete with OpenAI's Codex, expected to launch in coming months.

“Author of "Anthropic developing Claude Code web version to rival Codex" in TestingCatalog”

xAI is preparing to launch Imagine feature on Grok web with image-to-video conversion and Team accounts for enterprise workspace management.

“Author of "xAI readies Grok web with Imagine tool and Team profiles" in Testing Catalog”

OpenAI is preparing a ChatGPT Pro Lite tier at $100 per month to fill the pricing gap between its $20 Plus and $200 Pro plans and potentially support compute-intensive agentic capabilities.

“Author of "OpenAI Prepares New ChatGPT Pro Lite Tier Priced at $100 Monthly" in Testing Catalog”

Perplexity is testing a Messages integration for its Comet browser and developing a usage credits system to allow Pro users to purchase additional capacity after recent tier limit reductions.

“Author of "Perplexity Tests Messages Integration and Usage Credits" in Testing Catalog”

Microsoft is developing Copilot Advisors, a feature that lets users pit two AI personas with different stances against each other in structured debates on any topic.

“Author of "Microsoft Develops Copilot Advisors to Debate on Any Topic"”

Google is testing a NotebookLM integration for Opal that would enable users to reference their research notebooks as native data sources within automated no-code workflows.

“Author of "Google Test NotebookLM Integration for Opal Workflows" in Testing Catalog”

Meta AI is preparing to launch new models named Avocado with integrated Manus agent capabilities, voice support, scheduled task execution, and third-party app connectors via MCPs, with a possible February release.

“Author of "Meta AI Readies Avocado, Manus Agent, and OpenClaw Integration" in Testing Catalog”

Anthropic is preparing to launch Claude Sonnet 5 during Super Bowl week with competitive math performance and superior coding capabilities at lower cost than Opus, positioning itself to disrupt the consumer AI market.

“Author of "Anthropic is About to Drop Sonnet 5 During Super Bowl Week" in Testing Catalog”

Google is developing voice cloning capabilities for AI Studio, with a hidden 'Create Your Voice' feature found in the interface tied to Gemini 2.5 Flash, signaling preparation for a future rollout with advanced audio models.

“Author of "Google tests voice cloning on AI Studio powered by Gemini"”

OpenAI is developing shopping cart, merchant submission tools, and personalized chat features for ChatGPT to enable direct e-commerce transactions and expand the platform's utility in daily commerce tasks.

“Author of "OpenAI to add shopping cart and merchant tools to ChatGPT"”

Notion is expanding its Custom Agent platform with upcoming features including user-managed MCPs, Workers for custom integrations, computer use agents, and AI-powered writing assistance.

“Author of "Notion Working on Custom MCPs, Workers, and Computer Use"”

Anthropic is evolving Claude from a chat assistant into a comprehensive productivity agent with persistent knowledge bases, Cowork mode, expanded automation connectors, and voice interface capabilities.

“Author of "Anthropic Works on Knowledge Bases for Claude Cowork" in Testing Catalog”

Google's Stitch design agent is adding API key management and automatic PRD generation capabilities to streamline product workflows and integrate Gemini models for higher-resolution image generation.

“Author of "Google Stitch to Add API Keys and PRD Generation for PMs" in Testing Catalog”

Google is testing a new 'Auto Browse' capability for Gemini that would enable autonomous web browsing and tab management in Chrome, positioning it as a premium feature for Gemini Ultra users to compete with rival agentic AI platforms.

“Author of "Google Tests Gemini Auto Browse Tool for Chrome Users" in Testing Catalog”

xAI launched Grok Business ($30/seat/month) and Grok Enterprise plans enabling team and organizational deployment with Google Drive integration, data privacy protections, and advanced governance features.

“Author of "ICYMI: xAI launches Grok Business and Enterprise plans" in Testing Catalog”

Google is testing a new 'Lecture' format for NotebookLM's Audio Overviews that generates 30-minute AI-narrated lectures from source documents, with expanded voice options including British English arriving in 2026.

“Author of "Google Tests 30-Minute Audio Lectures on NotebookLM" in Testing Catalog”

Opera launches Neon, a paid agentic browser with global access at $19.90/month, offering AI agents that perform browser-based tasks and consolidated access to Gemini, GPT, and other leading models.

“Author of "Opera Launches Neon Browser Globally with Paid Early Access" in Testing Catalog”

Anthropic is expanding Claude with a new Agent mode featuring task delegation across five specialized sections (Research, Analyze, Write, Build, Do More) to position the assistant as an autonomous tool for structured workflows.

“Author of "Anthropic Preparing New Agentic Tasks Mode for Claude" in Testing Catalog”

Google is integrating NotebookLM directly into Gemini, allowing users to attach notebooks as live data sources for conversations with inline citations, rolling out to both free and paid accounts.

“Author of "Google Expands Gemini With NotebookLM Integration" in Testing Catalog”

Google is testing two new Gemini 3 models (Fierce Falcon and Ghost Falcon) on LM Arena, signaling an imminent public launch to compete with OpenAI's upcoming state-of-the-art models.

“Author of "Google tests Gemini 3 Flash model on LM Arena"”

OpenAI is testing Image-2 and Image-2-mini models that substantially improve visual quality and fidelity over Image-1, closing the gap with Google's Nano Banana 2, with expected launch coordinated alongside GPT-5.2.

“Author of "OpenAI Testing New Image-2 Models on LM Arena" in Testing Catalog”

Google is preparing to integrate its Opal Agent Builder into Gemini, enabling users to build and share AI-powered workflows directly within the platform without switching applications.

“Author of "Google Plans to Integrate Opal Agent Builder into Gemini" in Testing Catalog”

OpenAI is implementing a five-step app submission workflow for its ChatGPT Apps directory with enhanced compliance requirements, manual review, and structured validation processes ahead of public launch.

“Author of "Early Look at Upcoming App Submission Flow for ChatGPT" in Testing Catalog”

Google is developing a multi-agent system in Gemini for Enterprise that can generate and evaluate approximately 100 ranked ideas in a single 40-minute continuous run using tournament-style evaluation.

“Author of "Google to Enable Research Automation on Gemini Enterprise"”

Google will release Nano Banana Pro powered by Gemini 3 Pro next week, offering improved image and video generation capabilities for creative professionals targeting higher resolution output.

“Author of "Google to Release Nano Banana Pro Powered by Gemini 3 Pro Next Week" in Testing Catalog”

OpenAI is preparing a Group Chats feature for ChatGPT with custom system prompts and granular controls over AI response behavior, differentiating it from Microsoft's simpler Copilot group features.

“Author of "OpenAI Readies ChatGPT Group Chats With Custom Controls" in Testing Catalog”

xAI is developing Grok Code Remote, a cloud-based code execution feature to compete with OpenAI's Codex and Anthropic's Claude Code, with early access available through a December hackathon.

“Author of "xAI Works on Grok Code Remote to Rival OpenAI's Codex" in Testing Catalog”

OpenAI is preparing to release GPT-5.1 family models with enterprise availability targeted for November 24th, with early testing already underway on OpenRouter under the codename Polaris Alpha.

“Author of "You Can Now Test Upcoming GPT-5.1 on OpenRouter" in Testing Catalog”

OpenAI is preparing to launch GPT-5.1 Thinking model with multiple variants (Mini, Thinking, Codex-focused) optimized for advanced reasoning and multi-step tasks, timed to compete with Google's upcoming Gemini 3 Pro.

“Author of "OpenAI readies GPT-5.1 Thinking model ahead of Gemini 3 Pro" in Testing Catalog”

Google is set to release Gemini 3 Pro preview in November 2025 with a 1 million token context window, with broader availability expected in December.

“Author of "Google Plans to Release Gemini 3 Pro Preview in November" in Testing Catalog”

Google is preparing to release Nano Banana 2 (GEMPIX2), an updated image generation model for Gemini, with pre-release UI indicators suggesting a launch within the coming week.

“Author of "Google is Preparing Nano Banana 2 for the Upcoming Release" in Testing Catalog”

Perplexity is developing meeting assistant features including automatic recording, autonomous meeting joins, and automated recap distribution, positioning itself to compete with Google Meet and Microsoft in workplace AI.

“Author of "Perplexity Assistant will be able to join your meetings soon"”

OpenAI is testing GPT-5.1 Mini, a new model variant that showed improved image generation capabilities, suggesting an imminent launch possibly within November to compete with Google's upcoming Gemini 3.

“Author of "OpenAI Might Be Testing GPT-5.1-Mini Upgrade for ChatGPT"”

Google is preparing a public experiment for Genie 3 that will allow users to generate and explore interactive virtual worlds from text prompts.

“Author of "Google Prepares Genie 3 Public Experiment with AI-Generated Worlds"”

Google revamps AI Studio with democratized features including secret variable support, context-aware suggestions, and one-click Cloud Run deployment to enable rapid prototyping and aims for one million apps built by year's end.

“Author of "Google revamps AI Studio with new features for vibe coding"”

OpenAI is developing a shared prompts feature for ChatGPT team workspaces that enables organizations to create, standardize, and distribute reusable prompt templates with custom instructions and variables across collaborative environments.

“Author of "OpenAI Develops Shared ChatGPT Prompts Feature for Teams" in Testing Catalog”

Google's Veo 3.1 demonstrates significant improvements in visual fidelity, prompt adherence, and audio quality compared to Veo 3, positioning it as a competitive alternative to OpenAI's Sora 2 in generative video.

“Author of "First Real Samples of Veo 3.1 Generated Videos" in Testing Catalog”

xAI is preparing to launch advanced tooling for Grok that enables users to fetch and synthesize data from Gmail, Slack, Notion, and X, including real-time financial data via ticker symbols.

“Author of "xAI readies Grok tools support in Tasks for data fetching" in Testing Catalog”

Anthropic is preparing a major update introducing a hidden 'Skills Preview' feature that allows users to upload customizable instructions and automations to extend Claude's capabilities beyond current limitations.

“Author of "Anthropic Works on Customizable Skills for Claude" in Testing Catalog”

OpenAI is testing new alpha agent models with experimental variants labeled 'Agent with truncation' and 'Agent with prompt expansion' that suggest development of GPT-5-powered agents with improved context management strategies.

“Author of "OpenAI Tests ChatGPT Agent Upgrades Powered by New Alpha Models" in Testing Catalog”

OpenAI is preparing native checkout and order tracking capabilities in ChatGPT, along with Voice Mode UI improvements and parental controls for the web interface.

“Author of "OpenAI Readies Orders in ChatGPT for Native Checkout and Tracking" in Testing Catalog”

xAI launched Grok 4 Fast, a speed-optimized variant delivering up to 10x faster inference than standard Grok 4 by prioritizing response speed over output sophistication.

“Author of "xAI Launches Grok 4 Fast in Early Access Beta with Up to 10x Speed" in Testing Catalog”

Google's Jules coding agent ships PR comment replies, stacked diff viewer, and enhanced GitHub integration to streamline code review and project setup workflows.

“Author of "Google's Jules Ships Code Review Upgrades with PR Comments and More" in Testing Catalog”

Google is testing three new NotebookLM audio formats—Brief, Critique, and Debate—alongside expanded voice options, signaling an upcoming public rollout of enhanced audio customization features.

“Author of "Google Tests New NotebookLM Voices and Audio Formats Brief, Critique, Debate"”

Google is testing new Gemini modes including Agent Mode for autonomous task execution, Go for collaborative ideation, and Immersive View for visual explanations, positioning Gemini as a competitive agentic platform against ChatGPT.

“Author of "Google tests new Gemini modes including Agent, Go, and Immersive View"”

Google is integrating Deep Research capabilities from Gemini into NotebookLM, allowing users to pull sources from the web and Google Drive directly into their notebooks, while also preparing a new 'tutor' mode for interactive learning support.

“Author of "NotebookLM to Pull Sources from Web and Google Drive via Deep Research" in Testing Catalog”

Google is developing a projects feature for Gemini that enables users to organize files, manage project-specific instructions, and access research tools—following similar features from OpenAI's ChatGPT, Anthropic's Claude, and xAI's Grok.

“Author of "Google Develops Projects Feature for Gemini" in Testing Catalog”

Google is developing a code-named feature called Magic View for NotebookLM that appears to offer a new interactive way to visualize and interact with notebook data, while simultaneously expanding video overview and multilingual customization features.

“Author of "Google develops Magic View for NotebookLM as possible new way to visualize notebook data"”

Anthropic is internally testing Claude Opus 4.1 with improved problem-solving capabilities and undergoing safety validation with Neptune v4 before likely imminent release.

“Author of "Claude Opus 4.1 Likely in Internal Testing as Anthropic Prepares Safety Checks" in Testing Catalog”