We're excited to share that Anthropic has released Claude Opus 4.8, the most capable version of Claude to date. If you are an Anuma Starter or Pro user, you're already benefiting from this upgrade automatically behind the scenes.

Here's a breakdown of what's new and why it matters for you.

The headline numbers

• 1M Token context window (default)
• 128K Max output tokens
• 2.5× Faster output in Fast Mode
• 1,024 Token cache minimum

In practical terms, a 1 million token context window means Claude can retain an enormous amount of conversation history, documents, and project context in a single session, roughly the size of several novels. You’ll notice far fewer moments where the AI loses track of earlier discussions or important details.

What's actually improved

Smarter reasoning when it matters

Claude Opus 4.8 is better at deciding when a task requires deep reasoning versus a quick response. Simple questions get answered quickly, while more complex problems receive more thoughtful analysis automatically.

More reliable task execution

Previous versions could occasionally miss part of a multi step request. Opus 4.8 is significantly better at following through on every instruction, reducing back and forth and making complex workflows more dependable.

Faster responses when you need speed

The new Fast Mode delivers up to 2.5× higher output speed. That means quicker drafts, summaries, brainstorming sessions, and rapid answers when responsiveness matters most.

Stronger long context performance

Working through a long document, research session, or ongoing project over dozens of messages now feels much more stable. Opus 4.8 maintains focus better during extended agentic workflows with fewer derailments or forgotten details.

Anuma Users

You don’t need to change anything. Claude Opus 4.8 is now available in Anuma for Starter and Pro users.

The experience with Claude Opus 4.8 should feel noticeably sharper - faster when speed matters, more thoughtful when handling complex tasks, and more reliable during long conversations and multi step workflows.

Inside Anuma, Claude becomes even more powerful with unified memory across multiple AI models. Your conversations, saved knowledge, and context stay connected across workflows, so you spend less time repeating yourself and more time getting useful results.

Switch between Claude, GPT, Gemini, and other leading models while keeping the same shared memory and context throughout your workspace.

Using the Free plan? Upgrade to access the latest Claude model.

Source: Claude Opus 4.8 announcement via Anthropic