Anthropic releases Claude Opus 4.8 with dynamic workflows

#1 📝 Anthropic News

Introducing Claude Opus 4.8 - Anthropic released Claude Opus 4.8 today at the same price as Opus 4.7, adding user-selectable effort levels, Claude Code “dynamic workflows,” and a fast mode that runs 2.5× faster and is three times cheaper than on prior models. The company reports broad capability and alignment gains—Opus 4.8 is about four times less likely than Opus 4.7 to let code flaws pass unremarked, scored 84% on Online-Mind2Web, was the only model to complete every Super-Agent case end-to-end (beating prior Opus models and GPT-5.5), is the first to break 10% on the Legal Agent all-pass standard, and shows misalignment rates similar to Claude Mythos Preview while Genie users report 61% cheaper token cost versus Opus 4.7 for multimodal reasoning.

Also covered by: @v0, @There's An AI For That, @There's An AI For That, @Aravind Srinivas, @claire vo 🖤, building @chatprd, @Mike Krieger, @Claude, @Cognition, @Claire Vo, @Dan Shipper, @How I AI Podcast

#2 📝 Claude Code Blog

Introducing dynamic workflows in Claude Code - Product announcement on May 28, 2026 introducing dynamic workflows in Claude Code. The post announces a new workflow feature for Claude Code (details available on the linked article).

Also covered by: @Mike Krieger, @Claude, @Boris Cherny

#3 📝 Simon Willison

Claude Opus 4.8: “a modest but tangible improvement” - Anthropic released Claude Opus 4.8; Simon highlights a favorite note from the release announcement and links to his longer write-up and examples including pelicans generated with the new model. The entry links to Anthropic's announcement and contains an image card linking to the full post.

Also covered by: @v0, @There's An AI For That, @There's An AI For That, @Aravind Srinivas, @claire vo 🖤, building @chatprd, @Mike Krieger, @Claude, @Cognition, @Claire Vo, @Dan Shipper, @How I AI Podcast

#4 𝕏

Boris Cherny unveils Claude’s new dynamic workflows, which shrink major migrations and refactors from quarter-long engineering projects to just days or weeks.

Also covered by: @Mike Krieger, @Claude, @Boris Cherny

#5 📝 Anthropic News

Anthropic raises $65B in Series H funding at $965B post-money valuation - Anthropic raised $65 billion in Series H funding led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital at a $965 billion post‑money valuation and said its run‑rate revenue crossed $47 billion. The company said the round (including $15 billion of previously committed hyperscaler investments, $5 billion from Amazon) will fund safety and interpretability research and compute expansion—citing new agreements for up to 5 GW with Amazon, 5 GW of next‑generation TPU capacity with Google and Broadcom, GPU access via SpaceX Colossus 1 and 2—and that Claude is available on AWS, Google Cloud, and Microsoft Azure (with AWS as its primary cloud and training partner).

#6 📝 Simon Willison

Anthropic's $65B Series H announcement - Simon examines Anthropic's Series H announcement highlighting a claimed run-rate revenue of $47 billion and compares this with earlier stated run-rate figures and commentary. He argues the numbers are credible because they were disclosed in fundraising announcements and likely to be verified in future SEC filings.

#7 𝕏

Google Research unveiled Gemini for Science at I/O 2026 to accelerate scientific discovery and showcased AI-powered updates to Google Search, kicking off a new era of innovation.

#8 📝 OpenAI News

OpenAI’s Frontier Governance Framework - OpenAI publishes its Frontier Governance Framework outlining principles and approaches for governing frontier AI capabilities. The framework describes safety, research, and governance priorities for managing powerful AI systems.

#9 𝕏

Philipp Schmid announced GA release of Nano Banana 2 (Gemini 3.1 Flash Image) at $0.045/image and Nano Banana Pro (Gemini 3 Pro Image) at $0.134/image.

#10 𝕏

Qwen Qwen3.7-Max just hit #3 on ITbench-AA, a new agentic-style benchmark measuring model performance on real-world enterprise IT tasks.

#11 𝕏

Summary: NVIDIA AI adopts the Linux Foundation’s OpenMDW framework across its open model families to standardize licensing for models, code, documentation, and data, streamlining compliance and reducing friction for developers and enterprises.

#12 𝕏

xAI launched Grok Build 0.2.7 with new /usage and /login commands, shared terminals across subagents, and enhanced image understanding.

#13 𝕏

Teresa Torres breaks down how Lorikeet evolved from a command-line prototype into a dual-agent AI support system—Concierge for triage and Coach for live resolution—using “resolution in the loop” guardrails and a Trace Diagnosis Agent to know when to say “I can’t help.”

#14 𝕏

Guillermo Rauch launched the Vercel CLI as a self-updating, zero-dependency binary built on pkg to eliminate Node.js friction. He says it’s now central to the “cloud for agents,” with usage spiking via OpenClaw, Claude Code, and Codex.

#15 𝕏

Guillermo Rauch had his team use AI to spin up a static design and turn it into WebGPU shaders in under a day, highlighting how Next.js Nights will let you ship beautiful apps effortlessly by 2026.

#16 𝕏

Peter Yang built an AI-powered /slides skill that converts rough outlines into polished HTML decks in minutes, leveraging 12 slide formats, 3 templates, live charts, subtle animations, and auto-screenshot layout fixes.

#17 𝕏

v0 launched a new Figma integration that converts static designs into high-fidelity functional UI by importing layout, typography, components, icons, and images.

#18 📝 PromptLayer Blog

The 7 Best Prompt Management Tools in 2026 (Tested and Compared) - Overview and comparison of leading prompt management tools in 2026, explaining why prompt management becomes essential as prompts graduate from simple files to production assets. The article frames prompt management as necessary for versioning, deployment, testing, and monitoring.

#19 in

Peter Yang announces Google’s free 5-day AI Agents course on “vibe coding” agents—covering CLI builds, API/tool interop, memory, security and production deployment in just 1–2 hours a day. It drew 1.5 million learners last time.

#20 𝕏

Julien Chaumond, Co-founder and CTO @HuggingFace is bullish on data infrastructure after using Xet deduplication to clone 68 TB to a HuggingFace training bucket in just 1 minute 55 seconds despite only having a 4 TB local disk.

#21 𝕏

Santiago shares a Pika MCP Founder Kit Skills video featuring four agentic-coding marketing starters—brand identity/strategy/assets, high-converting product screenshots, a 15-sec product video, and a founder intro clip. He says these tools would have saved him months of work.

#22 𝕏

Santiago warns that binding your code to a single model API causes lock-in, maintenance headaches and overspending as pricing and capabilities shift.

#23 𝕏

bolt.new launched the Claude Marketplace, letting enterprise clients apply their existing Anthropic commitments to access Bolt and other integrated apps under the same budget.

#24 𝕏

Harrison Chase predicts every AI agent will need a dedicated code sandbox, which led to the launch of LangSmith Sandboxes GA to empower all types of agents—not just coding ones—to write and execute code for diverse tasks.

#25 𝕏

Cursor surfaces five emerging AI trends — developer acceleration, the economics of intelligence, the power-user gap, the rise of context, and the shift to automation — in its latest research report.