GenAI PM Brief Archive
AI product news, tools, research, and practical PM takeaways. Issues appear here after the subscriber-first publishing window.
Genie 4 rolls out on Gemini API for steerable 720p worlds
Read brief →Gemini 4 Neon rolls out; Genie 4 brings audio
Read brief →OpenAI releases GPT-6.2 Astra and Sol; Perplexity launches 45+ model Orchestrator API
Read brief →Perplexity Monitor tracks web changes for $1 per 1,000 runs
Read brief →Build Teammates on Perplexity with any frontier model
Read brief →Recap: Google Wallet connection in Gemini supports finance insights
Read brief →Cognition announces Devin Apps for private internal tools
Read brief →Google AI announces Gemini 4 Hydrogen; Claude Meetings enters beta
Read brief →Gemini Decisions API launches with 1,000-question multimodal scoring cap
Read brief →Google releases Gemini 4 Carbon for long-horizon agent workflows
Read brief →Anthropic upgrades Claude Fable and Mythos to 5.2
Read brief →Recap: Mural’s 30-minute GPT-Live-1 sessions cost about $3–$5
Read brief →Perplexity Teammate orchestrates email, calendars, briefs, and planning
Read brief →Vercel turns LLM output into Slack and Teams UI
Read brief →GPT‑6 Cyber announced; Gemini Omni 1.2 Flash API cheaper
Read brief →OpenAI introduces Images 2.6 with faster generation and GIFs
Read brief →Claude now auto-connects teammates to company plugins
Read brief →NVIDIA demos Cosmos 3 camera-controlled views from one image
Read brief →Vercel Labs’ Agent Desktop uses accessibility trees, not pixel matching
Read brief →Recap: PlanAssist made available to 12 more councils
Read brief →OpenAI announces Private Intelligence for sensitive business AI
Read brief →Mistral releases Le Chonk weights for coding and cybersecurity
Read brief →Google AI announces Gemini 4 Carbon cybersecurity model
Read brief →Google AI announces Genie 4 for real-time steerable worlds
Read brief →OpenAI introduces GPT-6.1 Terra for computer use; Cursor releases Agent SDK
Read brief →NVIDIA’s SkillSpector scans agent skills before installation
Read brief →Ollama adds macOS speech streaming and d3 decision models
Read brief →Cursor announces Computer Use for desktop apps; NVIDIA releases Nemotron 3.5 Bolt E4B 4GB
Read brief →OpenAI introduces Decisions API; Anthropic-managed Claude reaches Google Cloud GA
Read brief →Claude opens Duet preview; OpenAI models are coming to Google Cloud
Read brief →Perplexity Computer now runs fully locally on a Mac
Read brief →Meta details 30B open-weight Muse Glimmer for one consumer GPU
Read brief →OpenCode 1.19.3 generates reasoning variants from model options
Read brief →GPT-6 Astra Pro upgrade enables parallel agents
Read brief →Meta releases open-weight Muse Spark Open for coding
Read brief →Research@ recap covers Deep Review's document verification expansion
Read brief →Gemini 4 Argon is now globally available
Read brief →Cursor joins The X Subscription, with paid plans offering twice the Grok coding usage of equivalent Cursor plans
Read brief →TesterArmy e2e adds per-attempt branches to Chromium microVMs
Read brief →Open SWE’s model router cuts median task cost 64%
Read brief →Qwen-Image-2.1-Turbo gets open weights and 8-step 2K generation
Read brief →Gemini recap: sub-agents; Claude Dashboards, Motion are in beta
Read brief →OpenAI releases GPT-6 model to more ChatGPT users
Read brief →Mistral AI releases Large 4; DeepMind announces EmbeddingGemma 2
Read brief →Cognition introduces Dreaming: Devin refines cross-session memory overnight
Read brief →OpenAI announces ChatGPT visual ads and expanded measurement
Read brief →Perplexity Computer now generates interactive 3D learning visualizations
Read brief →Rollouts finds offending PRs; one click starts cloud agent
Read brief →pplx-decider-27b open-sourced; Decisions API priced at 4¢ per million input tokens
Read brief →Google AI announces Gemini 4 Argon's 1M-token output limit
Read brief →Anthropic announces faster, lower-cost Claude Sonnet 5.5
Read brief →Atlassian recap: Confluence cycle dropped to roughly six weeks
Read brief →Meta’s Muse agent runs in a cloud VM and uses a computer
Read brief →@finkd unveils Muse Realtime Avatar for real-time conversations
Read brief →Google AI launches Gemini 3.8 Flash and Flash-Lite TTS
Read brief →OpenAI News announces GPT-6 Sol and Luna
Read brief →Cognition announces Devin Cloud CLI and SSH workflows
Read brief →Warp’s Wilson turns Slack requests into tested PRs
Read brief →Qwen announces Qwen3.8-LiveTranslate, averaging 2.3-second lag across 60 languages
Read brief →OpenAI announces Astra for Law foundation
Read brief →Claude Cowork and chat merge into one Claude
Read brief →Google DeepMind introduces two Gemini 3.8 Live models
Read brief →Anthropic announces Claude Mods for Claude Code
Read brief →Brex’s Jim automates three recruiting workflows with OpenClaw
Read brief →fx.sh adds multi-model subagent orchestration
Read brief →Cognition announces Fusion for separate planning, execution models
Read brief →OpenAI introduces fully managed Agents API with Codex
Read brief →OpenAI previously shared a practical Defense Factory playbook
Read brief →How 1.5 engineers got Stripe Kai in two weeks
Read brief →How Kai’s three-layer data-query workflow works
Read brief →AIRA₃ coordinates agents through a forum and shared filesystem
Read brief →OpenAI releases GPT-6 Astra in Codex and API
Read brief →OpenAI announces GPT-6 Astra for computer tasks
Read brief →Google DeepMind announces Gemini 3.8 Flash and Flash Cyber
Read brief →Anthropic announces Claude Fable 5.1 and Mythos 5.1
Read brief →Meta’s Muse Code exits beta with agent SDK preview
Read brief →OpenAI builds for capabilities 2–3 months ahead
Read brief →OpenAI proposes ending Cursor’s direct model access November 12
Read brief →Google DeepMind rolls out Gemini Omni 1.1 Flash
Read brief →Anthropic previews Model Hardware Standard for agent-operated equipment
Read brief →Google AI announces context-aware Gemini 3.5 Transcribe
Read brief →NVIDIA previews Shadow engine recovery in Dynamo
Read brief →GPT-5.6 in Kiro advances developer price-performance
Read brief →Error Discovery Skill turns feedback into reusable evals
Read brief →Claude Security scans now run on Claude Mythos 5
Read brief →AI Extract launches with SQL-callable PDF extraction
Read brief →OpenAI announces Zero Data Retention for frontier models
Read brief →Claude adds Gmail sending and Drive file management
Read brief →Cursor releases Origin, its integrated code hosting platform
Read brief →Deepagents separates agent loops from filesystem-like backends
Read brief →Owning Your Intelligence: Portable Memory, Model-Agnostic Harnesses
Read brief →Perplexity releases Search SDK for agentic harnesses
Read brief →OpenAI previews Ultrafast: GPT-5.6 Sol up to 14× faster
Read brief →MAI-Thinking-1 is now available in Microsoft Foundry
Read brief →Commercial VLM recall falls below 35% after 50 pages
Read brief →Meta releases 30B Muse Glimmer for always-on local agents
Read brief →Linear Agent dynamically loads task-specific skills in production
Read brief →Claude Code sessions can now message each other
Read brief →Claude Fable 5 updates biology safeguards, reducing fallbacks
Read brief →Get each brief first
Subscribers receive new GenAI PM briefs before they enter the public archive.
Subscribe free