Xaira is an AI Protein & Enzyme Design company based in South San Francisco, United States.
| Date | Stage | Amount | Valuation | Lead investors |
|---|---|---|---|---|
| Apr 23, 2024 | Venture round | $1B | — | ARCH Venture Partners, Foresite Capital |
No articles ingested yet for Xaira. Once the hourly news pipeline is live, every article the classifier tags as mentioning this company appears here with its one-line AI summary and sentiment.
X-Design is Xaira's AI model for designing antibodies and other molecules, built on public and proprietary data. Its Vega class of models optimizes for human, potent, selective, cross-reactive and developable binders rather than binding alone, and generates designs zero-shot, without training on data from earlier campaigns. Xaira reported on Oct. 2, 2026 that it produced leads for its XA-1 oncology program and for XA-4, a GPCR target where display and immunization efforts had failed.
X-Cell is Xaira's virtual cell model for target identification, launched in March 2026. It predicts the outcomes of genetic perturbations in cells, helping the company choose targets worth pursuing. Xaira also released X-Atlas/Orion, a public Perturb-seq atlas, in June 2025.
X-Scientist, introduced on Oct. 21, 2026, is Xaira's agentic AI platform for drug discovery. It connects frontier language models to biological tools, internal datasets and Xaira's own models, including X-Cell and X-Design, to investigate mechanisms, nominate targets, design therapeutics and plan experiments. In a test on TGF-beta-treated liver cells, its nominated genetic perturbations achieved a hit rate comparable to human curation and were generated about ten times faster.
X-Automate, introduced on Oct. 23, 2026, is Xaira's laboratory automation system. It combines industrial workcells that express and purify thousands of protein samples a week, X-Bot, an in-house vision-guided six-axis robotic arm built from low-cost commercial hardware, and X-Scientist to coordinate workflows and safety checks. In a seven-day HPLC demonstration it autonomously retrieved, moved and analyzed 18 plates, 80% more than a modeled weekday loading schedule.
No recently verified openings are available right now. The company's careers page has the latest.
View all careers ↗