Anthropic’s Claude Agents Get ‘Dreaming’ Feature as Revenue Hits $30 Billion

Quick Facts

• Anthropic launched ‘dreaming’ capability for Claude agents, allowing them to process memories and improve performance between sessions
• Company’s annualized revenue run rate reached $30 billion, up from $9 billion at end of 2025, with 80-fold growth in Q1
• Anthropic is raising $50 billion at a valuation above $900 billion, surpassing OpenAI for the first time

Anthropic announced a ‘dreaming’ feature for its Claude AI agents at the Code with Claude developer conference on May 6. The capability allows agents to process recent events and identify useful memories to store for future tasks.

The feature runs between sessions to prune outdated notes, merge duplicates, and resolve contradictions in persistent memory files. Developers can manually trigger the process with the /dream command. The dreaming ability is currently in research preview and requires developer access requests.

‘Together, memory and dreaming form a robust memory system for self-improving agents,’ Anthropic wrote. ‘Dreaming surfaces patterns that a single agent can’t see on its own, including recurring mistakes, workflows that agents converge on, and preferences shared across a team.’

CEO Dario Amodei revealed explosive financial growth during the conference. The company planned for 10-fold growth but achieved 80-fold growth in the first quarter on an annualized basis. Anthropic’s annualized revenue run rate crossed $30 billion, up from $9 billion at the end of 2025.

Claude Code hit $1 billion in annualized revenue within six months of launch. The company now counts over 1,000 enterprise customers spending more than $1 million per year on Claude services, doubling since February.

Roughly 80% of Anthropic’s revenue comes from enterprise customers, with 40% of its top 50 clients being financial institutions, including Goldman Sachs, Visa and Citi. The company is now raising roughly $50 billion at a valuation above $900 billion, pushing it past OpenAI’s $852 billion valuation.

To address capacity constraints, Anthropic signed an agreement with SpaceX to use all compute capacity at the Colossus 1 data center in Memphis. This gives the company access to more than 300 megawatts of capacity and over 220,000 Nvidia GPUs.

Early testing shows promising results. Harvey used dreaming to help legal agents remember patterns, with completion rates rising about six times in tests. Netflix’s platform team used multi-agent orchestration to analyze logs from hundreds of builds and identify recurring issues.

Anthropic also announced multi-agent orchestration, allowing Managed Agents to break down complex tasks and assign them to sub-agents. The ‘outcomes’ feature helps AI agents focus on intent, with tests showing task success improvements of up to 10 points compared to standard prompts.

‘Software engineers are the ones who are fastest to adopt new technology,’ Amodei said. ‘It’s a foreshadowing of how things are going to work across the economy, and how the economy is going to be transformed by AI.’

Read more: Anthropic is letting Claude agents ‘dream’ so they don’t sleep on the job