Two narratives are running in parallel right now: the race to fund and scale frontier AI labs, and the quieter but steady work of making those models useful inside real developer pipelines. May 30 lands in an unusually busy stretch — Anthropic's valuation crossed $965 billion yesterday, Google's I/O keynote is still reverberating, and GitHub trending reflects a developer community that has moved firmly into the build phase.
Anthropic Raises $65B, Eclipses OpenAI in Valuation
Anthropic closed a $65 billion funding round led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital, pushing its valuation to $965 billion and surpassing OpenAI's most recent $852 billion mark for the first time. The company disclosed that its annualized run-rate revenue has crossed $47 billion — up from $10 billion at the end of 2025, roughly a 4.7x increase in under six months. Alongside the round, Anthropic confirmed that its upcoming Mythos model is nearing wide release, though no firm date was given.
Claude Opus 4.8: Sharper Coding, Honest Uncertainty
Released alongside the funding news, Claude Opus 4.8 raises the agentic coding benchmark from 64.3% to 69.2% and places deliberate weight on epistemic honesty — the model is far more likely to surface its own uncertainty rather than pushing forward with unsupported assumptions. Anthropic launched Dynamic Workflows in research preview for Claude Code, enabling longer autonomous task runs that previously required manual checkpointing, plus a Fast mode API preview offering up to 2.5x higher output throughput at premium pricing. Pricing holds at $5/million input tokens and $25/million output tokens, and the model is now available to GitHub Copilot Pro+, Business, and Enterprise users.
9to5Mac · GitHub Copilot changelog
Google I/O 2026: Gemini 3.5 Flash and WebMCP
Google's I/O keynote introduced Gemini Omni — aimed at multimodal creation from video, images, and text — alongside Gemini 3.5 Flash, which reportedly outperforms Gemini 3.1 Pro on most benchmarks while running four times faster, making it a strong fit for latency-sensitive agentic workflows. The most developer-relevant announcement is WebMCP, a proposed open standard that lets web applications expose structured JavaScript functions and HTML forms as machine-readable endpoints for browser-based AI agents, replacing fragile DOM scraping with a standardized interface. Google's Antigravity agent-first development platform also received major capability updates.
GitHub Trending: Documents, Voice Cloning, and Open CRM
microsoft/markitdown keeps climbing — 130K stars with nearly 1,900 added today — as the practical default for converting PDFs, Word docs, spreadsheets, and slide decks into clean Markdown for LLM pipelines. OpenBMB/VoxCPM jumped nearly 1,800 stars in a single day behind its VoxCPM2 model, which takes a tokenizer-free approach to multilingual text-to-speech and voice cloning. twentyhq/twenty (48K stars) continues to attract contributors as an open-source CRM built as an AI-native Salesforce alternative, with agent-driven data workflows designed into its architecture from the start.