Drop #14: Claude Opus 4.8 & a $65 Billion Vote of Confidence

On May 28, two announcements landed at the same time. Anthropic launched Claude Opus 4.8 — an upgrade to the Opus model family with better benchmarks, stronger honesty, and a new effort control feature in Cowork. The same day, the company closed a $65 billion Series H round at a $965 billion post-money valuation. Here is what changed and what it means for your Cowork workflows.

Claude Opus 4.8 Is Here

Opus 4.8 launched on May 28 as a direct upgrade to Opus 4.7. The most notable improvement is honesty: Anthropic's evaluations show Opus 4.8 is around four times less likely than Opus 4.7 to allow flaws in code it has written to pass unremarked. Early testers also report it is more reliable in long-running agentic tasks — it asks the right questions, catches its own mistakes, and pushes back when a plan is not sound. Computer use scores also improved significantly, with OSWorld-Verified rising to 84%.

Pricing is unchanged from Opus 4.7: $5 per million input tokens and $25 per million output tokens. Fast mode is now three times cheaper at $10 per million input and $50 per million output tokens. For Cowork users, Opus 4.8 is already the default — the upgrade happened automatically with no settings change needed.

Effort Control Lands in Cowork

Alongside Opus 4.8, Anthropic shipped effort control — a new setting in claude.ai and Cowork that lets you choose how much thinking Claude puts into each response. The control appears alongside the model selector and is available on all plans.

There are four levels. Low responds faster and conserves rate limits — good for quick lookups and simple formatting tasks. High (the default) balances quality and speed, using a similar number of tokens to Opus 4.7's default. Extra gives deeper reasoning and is recommended for complex tasks and long-running async workflows. Max uses the most tokens and produces the highest-quality output for the most demanding work.

Anthropic Raises $65 Billion

Also on May 28, Anthropic announced a $65 billion Series H funding round at a $965 billion post-money valuation, led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. The company's run-rate revenue crossed $47 billion earlier in May. The funding is directed toward safety and interpretability research, expanded compute capacity, and scaling the products and partnerships customers rely on. Anthropic CFO Krishna Rao noted the company is "working tirelessly to make tools like Claude Code and Cowork more helpful, more powerful, and more adaptable to their needs."

Tip of the Week: Match Effort to Task

Now that effort control is live in Cowork, use it intentionally. The default (High) is good for most tasks — but pushing it up or down at the right moments saves rate limits and gets better results.

❌ Weak

Leave effort on default for every task regardless of complexity.

✅ Strong

Scheduled task running overnight analyzing 50 reports?
Set effort to Extra. Unattended work benefits from deeper reasoning.

Quick reformatting of a file or a simple lookup?
Set effort to Low. Get the answer faster, preserve rate limits.

Everyday drafting, summarizing, and file work?
Stay on High (default). Best balance for most Cowork sessions.

The rule of thumb: the longer a task runs unattended and the higher the stakes, the higher the effort level you should pick.

Worth Reading This Week

Opus 4.8 and effort control together make Cowork more capable for complex, long-running work. The funding announcement signals that the pace of Anthropic investment is not slowing — more compute, more research, and more product updates ahead.

Until next time.

— The ClaudeCowork.com team