Profluent Bio
Revolutionizes genetic editing with AI-enhanced CRISPR precision.. [Contact for Pricing]
Last verified:
What is Profluent Bio?
Profluent Bio is an AI-first protein design company that uses frontier foundation models to author novel, functional proteins from scratch or inspired by natural scaffolds. The platform integrates advanced AI with an in-house wet lab to enable closed-loop discovery, aligning every design step with real-world use cases in medicine, agriculture, and biomanufacturing.
Key features include the ability to optimize multiple protein attributes simultaneously (affinity, stability, manufacturability), access greater sequence diversity beyond natural or patented proteins, and enable novel functionality through de novo protein design. The platform offers foundation models like ProGen3 for protein generation, ProseLM for structure-conditioned design, Profluent-E1 encoder models for fitness/structure prediction, and domain-specific models like Protein2PAM for CRISPR-Cas PAM prediction.
The company is for biotechnology companies, pharmaceutical partners, academic researchers, and industry teams working on therapeutics, gene editing tools, industrial enzymes, agricultural innovations, and biologics. Partnerships include discovery partnerships with tailored protein solutions and asset licensing for ready-made protein solutions.
Notable achievements include OpenCRISPR-1 (the world's first AI-designed gene editor), partnerships with Eli Lilly for AI-designed recombinases, Ensoma for base editors, Corteva for crop engineering, and RSRT for Rett syndrome treatments. The company has raised $150 million total funding and is backed by investors including Altimeter Capital, Bezos Expeditions, Spark Capital, and Insight Partners.
Profluent Bio pricing
Pricing model: Freemium
Profluent does not offer a public self-service pricing model. The company operates through partnerships and asset licensing. Discovery partnerships leverage advanced models to deliver tailored protein solutions with relevant wet-lab validation calibrated to partner needs. Asset licensing offers ready-made protein solutions designed to address common challenges with rapid iteration and further optimization. Some models are freely available: Profluent-E1 variants (150M, 300M, 600M parameters) are released freely for both research and commercial use. The Protein2PAM Server is an interactive platform freely accessible for academic and industry researchers worldwide. OpenCRISPR-1 is open-source. For partnerships, contact [email protected] or fill out the partnership form on their website. The Eli Lilly partnership included an upfront payment with Profluent eligible for up to $2.25 billion in development and commercial milestone payments plus tiered royalties.
Profluent Bio pros
- AI-first protein design with frontier foundation models
- Closed-loop discovery combining AI design with in-house wet lab validation
- Optimizes multiple attributes simultaneously (affinity, stability, manufacturability)
- Accesses untapped protein sequence spaces beyond natural proteins
- Enables de novo protein design for novel functionality
- Profluent-E1 encoder models freely available for research and commercial use
- OpenCRISPR-1 is open-source and successfully edits human genome
- Trained on Profluent Protein Atlas with billions of proteins
- Published in Nature and top AI conferences demonstrating scientific rigor
- Partnerships with major companies like Eli Lilly worth up to $2.25 billion
- Task-specific models like ProseLM for structure-conditioned design
- Protein2PAM server freely accessible for academic and industry researchers
- Three E1 variants (150M, 300M, 600M parameters) for different needs
- State-of-the-art fitness prediction without iterative laboratory screening
- Applications across therapeutics, agriculture, gene editing, and industrial enzymes
Profluent Bio cons
- No publicly disclosed pricing - partnership model only
- No free tier for full platform access - only select models open-source
- Primarily enterprise/partner-focused not consumer-facing
- Limited information on individual project costs or timelines
- Wet lab validation available only through partnerships
- No self-service web interface for protein design mentioned
- Requires direct contact for discovery partnerships and licensing
- Highly specialized tool requiring protein engineering expertise
Frequently asked questions about Profluent Bio
What is Profluent Bio?
Profluent Bio is an AI-first protein design company founded in 2022 and headquartered in Emeryville, CA. The company develops deep generative models and frontier foundation models to design and validate novel, functional proteins. Profluent turns biology into something programmable, creating solutions that transform medicine, agriculture, and beyond by authoring new biology grounded in nature with AI as an interpreter.
What is OpenCRISPR-1?
OpenCRISPR-1 is the world's first AI-designed gene editor, demonstrating Profluent's authorship capability in action. It was successfully used to edit the human genome and is available as open-source. The design was published in Nature in July 2025 with new results showing OpenCRISPR-1 glimpses into a generalizable platform for genome editing.
What is Profluent-E1?
Profluent-E1 is a family of retrieval-augmented protein language models (RA-PLMs) that enhances the PLM training paradigm by explicitly providing relevant evolutionary context through homologous protein sequences during training and inference. Unlike ProGen3 which is generative, E1 is an encoder model learning representations for downstream tasks. Three variants exist with 150M, 300M, and 600M parameters, released freely for research and commercial use for fitness prediction, structure prediction, and representation learning.
What is the Profluent Protein Atlas?
The Profluent Protein Atlas (PPA) is the most extensive and carefully curated protein dataset available to date, containing billions of proteins and trillions of amino acids. Profluent's Protein Language Models are trained on this dataset, implicitly learning evolutionary patterns shaped by natural selection over billions of years.
How does Profluent's platform differ from traditional protein design?
Profluent's foundation models push protein design beyond the limitations of random discovery. The platform integrates frontier foundation models with an in-house wet lab for closed-loop discovery, enabling alignment of every design step with use-cases that matter. Unlike traditional methods constrained by natural or synthetic biases, Profluent can extrapolate into new protein spaces, optimize multiple attributes simultaneously, and create de novo proteins with novel functionality.
What partnership options are available?
Profluent offers two main partnership types: Discovery partnerships leverage advanced models to deliver tailored protein solutions with relevant wet-lab validation calibrated to partner needs, and Asset licensing explores ready-made protein solutions designed to address common challenges offering rapid iteration and further optimization. Partnerships can involve enhancing existing scaffolds or designing new proteins from scratch.
What modalities does Profluent work with?
Profluent works with multiple protein modalities including antibodies (with precise control over antibody design enabling better biologics with epitope specificity and bespoke therapeutics for difficult targets), CRISPR-Cas systems for genome editing, base editors for gene therapies, recombinases for kilobase-scale DNA editing, and industrial enzymes. The platform can design antibodies for cryptic targets, multi-specifics, and ensure manufacturability.
Who funds Profluent Bio?
Profluent has raised $150 million total funding. The $106 million financing round was co-led by Altimeter Capital and Bezos Expeditions, with participation from existing investors including Spark Capital, Insight Partners, and Air Street Capital. Other investors include AIX Ventures and Convergent Ventures.
What are Profluent's key publications?
Key publications include 'Design of highly functional genome editors by modelling CRISPR–Cas sequences' published in Nature (July 2025), 'Large Language Models Generate Functional Protein Sequences Across Diverse Families' (January 2023), 'Deep Extrapolation for Attribute-Enhanced Generation' (July 2021), 'Adapting protein language models for structure-conditioned design' (August 2024), and 'Engineering of CRISPR-Cas PAM recognition using deep learning of vast evolutionary data' (January 2025).
How can I contact Profluent Bio?
For partnerships and general inquiries, contact [email protected] or fill out the partnership form on their website. For press inquiries, contact [email protected]. For partnership-specific contact, Cas Smith can be reached at [email protected]. The company is headquartered in Emeryville, California and can also be followed on LinkedIn, X, and YouTube.