Three proprietary models: AINN-P1, AINN-Basemodel, AINN-Peptide are available as secure on-demand APIs that integrate directly into existing discovery pipelines
SAN FRANCISCO , CA, UNITED STATES, June 12, 2026 /EINPresswire.com/ — Ainnocence, Inc., a next-generation AI drug discovery company, announces the production launch of its core protein-AI models as secure, on-demand APIs. Three of the company’s proprietary models, the AINN-P1 protein foundation model, the AINN-Basemodel antibody-antigen binding predictor and the AINN-Peptide therapeutic-property platform are now live, allowing pharmaceutical and biotechnology partners to integrate Ainnocence’s predictions directly into their own discovery pipelines.
From Benchmarked Models to Production Endpoints
Over the past year Ainnocence has reported a series of validated results: the AINN-P1 protein foundation model posted state-of-the-art Spearman correlations among single-sequence models on the ProteinGym leaderboard, and the Ainnocence Peptide AI Platform screened more than 3.2 million peptide variants at greater than 90% activity-classification accuracy. Today’s announcement closes the gap between benchmark and bench: those same models are now available on demand. What previously required for a dedicated research collaboration is now a single API call that returns results in seconds.
Three Models, One Integration Pattern
All three models share one consistent interface, so a partner who integrates one can adopt the others with little additional effort. Each accepts a protein or peptide sequence (or, for antibodies, a set of chains) and returns clear, structured predictions ready to drop into downstream analysis.
AINN-P1, Protein foundation model. Trained on UniRef sequences and served in three sizes (167M, 500M, and 1B parameters) from a single endpoint. It returns rich high-dimensional sequence embeddings, the universal feature basis for downstream property prediction with additional research-stage capabilities for sequence scoring and generation.
AINN-Basemodel, Antibody-antigen binding affinity. A graph-neural-network ensemble that scores how strongly an antibody is predicted to bind a target antigen on a 0-to-1 scale. A dedicated batch endpoint ranks up to roughly a thousand antibody-antigen pairs in a single request, turning candidate triage into a one-call operation.
AINN-Peptide, Therapeutic Property Platform. A dual-module system that, in one call, classifies a peptide and profiles it across 13 therapeutic activities from antimicrobial and anticancer to antiviral and immunomodulant giving discovery teams a comprehensive functional readout per sequence.
Production Performance and Scale
Delivered as production APIs, the models return single-sequence predictions in well under a second many in tens of milliseconds and sustain high-throughput batch screening at the multi-million-variant scale demonstrated in Ainnocence’s published benchmark studies. Capacity scales elastically with demand, so partners can move from a single query to a large screening campaign without changing their integration.
Privacy by Design: The Models Never Leave
Protecting model intellectual property was a first-class design requirement. Partners interact with the models exclusively as service sequences go in; predictions come out, so the models themselves are never exposed, copied, or redistributed. Partners gain the full benefit of Ainnocence’s foundation models within their own workflows, while the underlying science remains with Ainnocence’s protected IP.
Reliable Access for Every Partner
Every partner receives their own API credentials with usage tiers matched to their needs. This ensures fair, reliable access for all partners, protects performance during high-volume screening, and gives each team a clear, predictable allocation that can grow with the partnership.
Built for Integration
The APIs are built to fit the way discovery teams already work, integrating cleanly with existing computational pipelines, electronic lab notebooks, and cloud workflows. Adding Ainnocence’s predictions to an automated screening or ranking workflow takes only a few lines of code.
“A model is only valuable if teams can use it. By delivering AINN-P1 and our peptide platform through secure APIs, partners can access the predictive power of our foundation models directly from their own pipelines without exposing the underlying science. What once took months to integrate now takes an afternoon.” – Dr. Lurong Pan, PhD, Founder and CEO of Ainnocence.
About Ainnocence
Ainnocence is a next-generation AI drug discovery company founded in 2021 and headquartered in Mountain View, California. The company’s platform encompasses a generative AI engine able to screen up to 10 billion molecules within hours to accelerate drug discovery, CarbonAI® for small-molecule and PROTAC design, SentinusAI® for antibody engineering, CellulaAI™ for cell therapy, NatmolAI™ for natural product discovery, and additional AI engines spanning target assessment and biosynthesis. Ainnocence partners with pharmaceutical and biotechnology organizations to compress discovery timelines and improve the quality of candidates advancing into development. For more information, visit ainnocence.com.
Lurong Pan, PhD
Ainnocence
+1 205-249-7424
lurong.pan@ainnocence.com
Visit us on social media:
LinkedIn
YouTube
X
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
Media gallery
