Verified Teams Now Access Mythos 5.1, Opus 5 for Biology Research
The Life Sciences Verification Program (LSVP) launched September 17, 2026, giving verified research teams access to Mythos 5.1, Opus 5, and Sonnet 5 with safeguards tuned for biology work. Two grant types — Standard Use (team-wide, annual) and High-risk Use (per-project, 6-month) — cover everything from daily R&D to dual-use research. Applications are open now for Enterprise and Team plans.
As of September 17, 2026, life science teams on Claude Enterprise and Team plans can apply for verified access that unlocks biology tasks previously blocked in the generally-available Fable models. The Life Sciences Verification Program (LSVP) grants access to Mythos 5.1, Opus 5, and Sonnet 5 with classifiers tuned for biology — covering drug discovery, research biology, clinical development, and manufacturing workflows.
Dozens of organizations joined through early access. The program is now open to the broader life science community, with Anthropic expecting to enroll hundreds of organizations within the first week.
Two grant types, one decision to make
Standard Use covers most life science work. It grants an entire team access to Mythos 5.1, Opus 5, and Sonnet 5, renewed once a year. It's built for daily, diverse workloads: basic science, R&D, supply chain, clinical development, quality assurance, regulatory affairs, and investing. New to LSVP? Start here.
High-risk Use is an add-on for work that stays blocked under Standard Use. It applies to a single research project — not a full team — and renews every six months. A researcher doing dual-use work might hold one Standard Use grant for daily tasks and one High-risk Use grant for a specific project, like characterizing how one family of viral vectors is recognized by human immune pathways.
High-risk grants for Opus 5 and Sonnet 5 are available today. High-risk grants for Mythos are limited to a small set of entities pending additional US government vetting.
Verification and how monitoring works
Every applicant is reviewed for research credentials, security standards, and ethical research oversight. Once verified, organizations describe their intended use cases — at the level of detail you'd find in a job listing, without sensitive IP — and Anthropic monitors LSVP traffic against those stated scopes.
The key design shift: LSVP monitoring is offline, not real-time blocking. Real-time blocking rejects requests one by one. Offline monitoring looks at patterns across sessions, catching misuse that is deliberately spread out to evade detection while letting legitimate work proceed with fewer interruptions.
Flagged activity is retained for 30 days and triggers notification to organization admins within pre-agreed timeframes. That data is strictly compartmentalized — it cannot be used for model training and is not accessible to Anthropic's life sciences research teams. All other safeguards, such as cyber classifiers, remain active under LSVP grants.
Where LSVP works today — and where it doesn't yet
LSVP is available through the API console, Claude Science, Claude.ai, and Claude Code for Enterprise and Team plans. Three categories are not yet supported: individual Pro and Max plans (coming later), third-party platforms, and BAA-enabled organizations (PHI customers should use a separate non-HIPAA org).
In Claude.ai and Claude Code, your default grant applies automatically — there is no in-product grant switcher yet. Switching between grants is possible in the API and Claude Science, or in Claude Code when using API authentication. Most users who only need Standard Use will never notice the limitation.
What you can do now: If your organization qualifies — academic lab, startup, pharma company, or otherwise — apply for LSVP. Request Standard Use for your team's daily work. Add a High-risk Use grant only for specific projects that need it. Write your use-case description like a job listing: what the team does, not how it does it.
Key numbers to remember: Standard Use renews annually. High-risk Use renews every 6 months. Flagged data is retained for 30 days. Mythos 5.1, Opus 5, and Sonnet 5 are on Standard Use today. High-risk Mythos access is restricted pending US government collaboration.