Anthropic's Claude Science bets on AI-led drug discovery

The News

Anthropic has moved deeper into laboratory research with Claude Science, an "AI workbench for scientists" that the company released in beta on 30 June 2026. The tool stitches together the fragmented software, databases and computing resources that researchers juggle daily, and it is available to subscribers on Claude Pro, Max, Team and Enterprise plans.

Rather than a single chatbot, Claude Science ships with more than 60 curated skills spanning genomics, single-cell analysis, proteomics, structural biology and cheminformatics. It can render 3D protein structures, genome tracks and chemical diagrams natively, scale compute from a laptop to an HPC cluster or on-demand GPUs, and produce reproducible artefacts complete with code history and plain-language notes. A built-in reviewer agent checks citations and calculations before results ship.

The launch leans heavily on drug discovery. Anthropic has wired in NVIDIA's BioNeMo Agent Toolkit, giving access to biology models such as Evo 2, Boltz-2 and OpenFold3, and partnered with compute provider Modal. To seed adoption, it is offering up to $30,000 in credits across 50 "AI for Science" projects, with Modal adding up to $2,000 of compute per project. Applications close on 15 July 2026, and selected work runs from 1 September to 1 December 2026. Early users include biotech firm Manifold Bio and researchers at the Allen Institute and UCSF, where epidemiologist Stephen Francis said the tool "dramatically accelerated the analysis".

Why It Matters

Frontier labs have spent two years proving their models can write code and draft emails. The harder, more valuable frontier is science, where a genuine acceleration of discovery reshapes economies rather than workflows. By packaging domain skills, compute orchestration and an audit trail into one product, Anthropic is betting that AI's next commercial moat sits at the laboratory bench, not the inbox.

The move also sharpens a rivalry. Google DeepMind's AlphaFold reset expectations for computational biology when it arrived in 2020 and later earned a share of Nobel recognition. Anthropic's answer is less a single breakthrough model than an operating environment, an attempt to own the daily grind of research rather than one celebrated result. If it works, the company captures recurring subscription revenue from an audience that has historically bolted its own tools together for free.

Indian Angle

For India, the timing is pointed. The country runs one of the world's largest pharmaceutical industries by volume, and firms such as Sun Pharma, Dr Reddy's and Cipla have been edging from generics towards novel-molecule research where computational tools increasingly decide the pace. A subscription-priced workbench that folds in protein modelling and cheminformatics lowers the capital barrier that has long favoured better-funded Western labs.

It also lands amid policy tailwinds. India's BioE3 policy, cleared in 2024, explicitly targets biomanufacturing and AI-driven biology, while the Department of Biotechnology has been funding computational research at the IITs, IISc and CSIR laboratories. Home-grown players such as Aganitha and Elucidata already sell AI-for-science tooling, so Anthropic's entry is both validation of their thesis and fresh competitive pressure.

The catch is cost. Credits and subscriptions are billed in US dollars, and a soft rupee makes sustained access expensive for smaller Indian labs and academic groups once the free credits lapse. Whether Indian institutions can budget for always-on frontier compute, or lean on domestic alternatives, will shape how far these tools reach the country's deep bioinformatics talent pool.

FAQ

When is Claude Science available?

It entered beta on 30 June 2026 for subscribers on Claude Pro, Max, Team and Enterprise plans. The linked AI for Science grant programme accepts applications until 15 July 2026, with funded projects running from 1 September to 1 December 2026.

How is it different from a normal Claude chat?

Claude Science bundles more than 60 research-specific skills, renders 3D molecular and genomic structures, orchestrates compute from laptops to HPC clusters, and adds a reviewer agent that checks citations and calculations, capabilities a standard chat interface does not provide.

What does it mean for Indian pharma?

It lowers the cost of computational drug discovery for firms and academics moving beyond generics, though dollar-denominated pricing and domestic rivals such as Aganitha and Elucidata will heavily influence how quickly it is adopted.

Where can I read the original announcement?

The Verge first reported the launch, while Anthropic's own newsroom carries the full product detail and the terms of its AI for Science grant programme.

This story was reported by The Verge. Read the full original coverage at The Verge.