Anthropic's AI research workbench that runs computational science end to end — analysis, visualization, and reproducible outputs in one place.
Last verified · 2026-06-30 · by Moe Ameen
Claude Science is an AI workbench for scientists from Anthropic, released in beta on June 30, 2026. Instead of bouncing between databases, code environments, and analysis tools, a researcher works in one place: a coordinating Claude agent acts like a project lead, with access to 60+ curated databases (UniProt, PDB, Ensembl, ClinVar, ChEMBL, GEO and others) and prebuilt skills for fields like genomics, single-cell analysis, proteomics, structural biology, and cheminformatics.
The architecture is its defining idea. A generalist agent plans the work and can spin up sub-agents or hand tasks to specialist "expert" agents a user has built for their own research. A separate reviewer agent then checks citations and calculations, flagging and correcting errors before anything reaches a manuscript. The app natively renders scientific artifacts — protein structures, molecules, sequence alignments, genomic tracks, PDFs — and keeps full lineage from raw data to publication-ready figures, so results stay reproducible.
Importantly, Anthropic was explicit that Claude Science is not a new model and not a specialized biology model. It runs the same Claude models already available to everyone, including Claude Opus 4.8, with no special access or gating — the product is the workflow, not new weights. The app is in beta on macOS and Linux for Pro, Max, Team, and Enterprise plans, with a discounted Team plan for academic labs and nonprofit research organizations. Treat any availability or plan detail as a snapshot and check Anthropic's page before relying on it.
One thing to keep straight: Claude Science is built to do and document research — run analyses, manage compute, draft manuscripts, keep figures honest. It is not a content or marketing tool. Turning a finding into a public explainer video, a carousel, or a newsletter is a separate job downstream of the science.
Claude Science gets you to a defensible result — a figure, a finding, a reviewed manuscript. What it does not do is turn that result into something a non-scientist audience will actually watch, read, and share. That translation from research artifact to public content is where Kompozy picks up. Take the headline finding from a Claude Science project — a key chart, the one-line result, the "why it matters" — and bring it into Kompozy as a source. From that single input the engine fans out a science-explainer set: a Persona or avatar video where a presenter walks through the result, a Carousel that rebuilds the figure as branded slides, a Blog Article that explains the method in plain language, an Email Newsletter for your list, and platform-native posts for LinkedIn, X, and the rest — all written in your lab or brand's voice through a Persona Brief, then captioned, scheduled, and published across nine platforms from one queue.
The concrete workflow for a research lab, biotech, or science communicator: the science stays in Claude Science, where reproducibility and citation-checking belong, and the communications stay in Kompozy, where media and distribution live. You are not asking a research workbench to do marketing, and you are not pasting figures into a dozen tools by hand. One reviewed result becomes a week of on-brand content — an explainer short, a figure carousel, a blog write-up, and a newsletter — without a separate design or scheduling tool in the chain.
Claude Science is an AI research workbench from Anthropic, launched in beta on June 30, 2026, that gives scientists one environment for computational research — a coordinating Claude agent with 60+ databases and prebuilt skills, sub-agents and custom expert agents, a reviewer agent that checks citations and calculations, native rendering of scientific artifacts, and reproducible outputs. It is available on macOS and Linux for Pro, Max, Team, and Enterprise plans.
No. Anthropic was explicit that Claude Science is not a new model and not a specialized biology model. It runs the same Claude models already available to everyone, including Claude Opus 4.8, with no special access or gating. The product is the research workflow built around those models, not new weights.
Research scientists doing computational work — genomics, single-cell analysis, proteomics, structural biology, cheminformatics — including those without a deep coding background. It is built to consolidate databases, compute, analysis, and manuscript drafting into one reproducible workspace, not to produce marketing or social content.
No. It is a research and analysis tool — it produces figures, analyses, and manuscript drafts, not media or published posts. To turn a finding into an explainer video, a figure carousel, a blog, or a newsletter across platforms, you pair it with a content engine like Kompozy that handles the media, branding, and multi-platform publishing.
It is in beta for Anthropic's Pro, Max, Team, and Enterprise plans, with a discounted Team plan for academic labs and nonprofit research organizations; Team and Enterprise admins must enable it. Anthropic did not publish standalone pricing at launch, so check its official page for current plan and access details.