Anthropic’s Claude Science targets scientists’ workflows as rivals race on raw AI power

Anthropic’s Claude Science targets scientists’ workflows as rivals race on raw AI power
Anthropic is pushing into scientific research labs with Claude Science, a new AI “workbench for scientists” that aims to reorganize how computational biology and drug discovery are done, even as rival labs race ahead on ever-more powerful models.

Early June 30: Anthropic’s pitch to scientists

Anthropic formally introduced Claude Science as “an AI workbench for scientists,” describing it as an app that integrates commonly used research tools, produces auditable artifacts, and offers flexible computing access in one environment. The company says the platform tackles the tedium of juggling PubMed, Jupyter, R, cluster terminals, and bespoke data pipelines by bringing “fragmented tools into a single research environment” that covers all stages of work, from literature analysis to multi‑step experiments and manuscript drafting.

Anthropic emphasizes that Claude Science is not a new underlying model but a coordinating agent with access to “over 60 curated skills and connectors” tailored to genomics, single‑cell, proteomics, structural biology, and cheminformatics, supplemented by a reviewer agent that checks citations and calculations and flags errors.

Media reaction: workflow over model power

Soon after the announcement, tech outlets framed the launch as a strategic move to own scientists’ workflows rather than compete on raw model capability. TechCrunch noted that Claude Science “is ‘not a new AI model and not a more capable model for biology’… It runs the same Claude models already available to everyone today (including Claude Opus 4.8), with no special access and no gating,” positioning it as a unified environment that spares researchers “the hassle of bouncing between databases, pipelines, and tools.”

The Next Web similarly described Claude Science as an app that “pulls a researcher’s scattered tools into one place and lets AI agents run large parts of the work,” calling it Anthropic’s “biggest push yet into the lab.” The outlet highlighted a reviewer agent that inspects outputs “for untraceable figures and references that do not match the code,” aiming to catch AI‑generated mistakes before humans do.

The Verge captured the lab‑focused rebrand with the line “Anthropic launched Claude Science in beta today, calling it an ‘AI workbench for scientists.’ The tool pulls fragmented tools and datasets into one environment, and generates figures and visuals like 3D protein structures,” while stressing that “Claude Science is not a new AI model.”

The Financial Times, meanwhile, tied the launch directly to industry economics, headlining that “Anthropic launches Claude Science in push for pharma revenue,” and underscoring use cases in 3D protein rendering and drug discovery.

Competitive backdrop: benchmarks for biological reasoning

Anthropic’s workflow bet lands amid intensifying competition on scientific AI benchmarks. On the same day, OpenAI president Greg Brockman promoted GeneBench‑Pro, a new test for judgment‑heavy computational biology, saying it evaluates whether models can handle “the kind of judgment‑heavy analysis that real‑world computational biology requires,” with problems that would take human experts “around 20–40 hours to complete,” and touting GPT‑5.6 Sol as “a big step forward.”

Together, these moves highlight two diverging strategies: Anthropic trying to win scientists by owning the lab workflow, and rivals like OpenAI pushing the frontier on model‑level biological reasoning.

Continue reading https://foxvector.com/stories/019f60ec-c5b5-0e2b-73f7-2fbb782d232e