Profluent

Profluent's April 22, 2024 OpenCRISPR report describes functional testing of 48 AI-generated Cas9-like sequences in human cells and reports OpenCRISPR-1 on-target editing of 55.7% versus 48.3% for SpCas9, with off-target editing of 0.32% versus 6.1%.

Profluent has active partnerships across therapeutics, agriculture and protein engineering, including a 2026 Lilly deal worth up to $2.25B, but broad downstream commercial deployment is still early.

Profluent has active partnerships across therapeutics, agriculture and protein engineering, including a 2026 Lilly deal worth up to $2.25B, but broad downstream commercial deployment is still early.

Profluent created OpenCRISPR-1, the first open-source AI-generated gene editor; its protein language models generated millions of CRISPR-like proteins not found in nature, and OpenCRISPR-1 demonstrated comparable on-target editing efficiency with higher specificity than SpCas9 across tested targets.

Profluent develops frontier protein language models and AI-designed gene editors, including OpenCRISPR and ProGen3, with de novo protein design capabilities validated experimentally.

Profluent develops frontier protein language models and AI-designed gene editors, including OpenCRISPR and ProGen3, with de novo protein design capabilities validated experimentally.

In September 2026 Profluent announced a named BioMarin collaboration combining its protein-design AI with BioMarin enzyme replacement therapy expertise to develop candidates for genetically defined conditions.

Profluent combines foundation-model research, in-house wet-lab validation and strategic partnerships across therapeutics, agriculture and protein engineering.

Profluent combines foundation-model research, in-house wet-lab validation and strategic partnerships across therapeutics, agriculture and protein engineering.

Profluent's OpenCRISPR repository releases an AI-designed gene-editor protein sequence and compatible guide RNA with protocols and compatibility with canonical Cas9 guide RNAs; commercial therapeutic use requires a license agreement.

Foundation-model and protein-design skills transfer broadly within biotech AI and adjacent biological engineering, while gene-editing and wet-lab specialization limit full cross-market portability.

Foundation-model and protein-design skills transfer broadly within biotech AI and adjacent biological engineering, while gene-editing and wet-lab specialization limit full cross-market portability.