Anthropic launches Claude Science, an AI workspace for scientists, in beta

Anthropic, the developer of the Claude family of AI models, unveiled Claude Science on June 30, 2026. The app pulls a researcher's scattered tools into a single workspace and lets AI agents take over a significant share of scientific work. The product is available as an open beta, and the company calls it its biggest move into science to date.

What Claude Science can do

Claude Science is designed as a single desktop for researchers. Instead of switching between dozens of separate tools, from literature search to data analysis, scientists get one environment where AI agents built on Claude models can handle a significant share of routine tasks on their own.

• Launch date: June 30, 2026
• Product status: open beta
• Claude Science brings a researcher's scattered tools into one app
• Claude AI agents take over a significant share of research work
• Anthropic calls it its biggest step into the lab to date

Why it matters for researchers

Scientific work is full of routine but labor-intensive tasks: searching the literature, organizing data, running first-pass analysis of results. These are exactly the tasks Anthropic wants researchers to delegate to AI agents inside Claude Science, freeing up more room for the substantive part of the work. By calling the launch its biggest move into science, Anthropic signals that it sees labs as one of the key use cases for agentic AI, on par with other professional fields.

Context: agentic AI moves beyond the chat window

Leading AI labs have been steadily turning their models from simple chat interfaces into agents that can carry out multistep tasks on their own rather than just answer individual questions. Claude Science follows the same logic: instead of only responding to a scientist's queries, the system is meant to act inside their workflow. It collects data, drafts analyses, and takes over other steps that previously required the researcher to do the work by hand from start to finish.

Choosing science as the testing ground for such a product is no accident. Research consists of many clearly defined stages where it is easy to measure whether an agent completed a task or not. That makes the lab a convenient proving ground for testing how far AI agent autonomy can be pushed before similar solutions move into other professional fields.

What it means

The launch of Claude Science confirms that, for Anthropic, AI agents are not just a chatbot with extra features but a tool that plugs into a professional's real workflow and takes over part of their day-to-day tasks. If the beta lives up to expectations, science could become one of the first fields where agentic AI noticeably speeds up the routine side of research.

FAQ

What is Claude Science?

It is an app from Anthropic that brings a researcher's scattered tools into one place and lets AI agents built on Claude models handle a significant share of scientific work.

When did Claude Science become available?

Anthropic announced the launch on June 30, 2026; the product is available as an open beta.