dsm-firmenich Partners With Boltz to Accelerate AI-Powered Flavor and Fragrance Ingredient Discovery

dsm-firmenich has partnered with Boltz PBC to apply advanced artificial intelligence models to the discovery of new flavor and fragrance ingredients, combining Boltz's biomolecular AI technology with dsm-firmenich's extensive taste, odor and receptor data.

Announced on September 14, 2026, the partnership is designed to accelerate the discovery and design of new molecules for scent and flavor applications. The companies will develop exclusive models that combine AI-driven molecular modeling with decades of dsm-firmenich expertise in sensory science and receptor biology.

dsm-firmenich Brings AI Into Ingredient Discovery

Artificial intelligence is increasingly becoming an important tool in chemical and ingredient discovery. Rather than relying only on conventional laboratory experimentation, AI models can help scientists analyze large molecular datasets and identify promising candidates for further research.

dsm-firmenich has already been using AI across its businesses and says it created the world's first AI-created flavor in 2020. The company's latest collaboration with Boltz takes this approach further by applying biomolecular AI models specifically to taste and odor receptors.

The objective is to give perfumers, flavorists and scientists new tools for exploring molecular possibilities and identifying promising ingredients more efficiently.

Partnership Combines Boltz AI With dsm-firmenich's Scientific Data

The collaboration brings together two different areas of expertise.

Boltz develops AI models for biomolecular modeling, including technologies for molecular structure prediction, affinity estimation and generative molecular design.

dsm-firmenich, meanwhile, has decades of proprietary knowledge and data related to sensory science, taste and odor receptors, chemistry and ingredient development.

By combining these capabilities, the companies aim to build models that can predict and design molecular structures while also helping scientists understand how molecules interact with specialized proteins involved in taste and smell.

AI Models to Study Taste and Odor Receptors

A central part of the partnership is the use of AI to study taste and odor receptors.

Receptors are specialized proteins that detect chemical signals. In the human sensory system, these receptors help translate molecules encountered through food, beverages and fragrances into signals that the brain can interpret.

Understanding how molecules interact with these receptors can help scientists investigate why particular compounds produce specific taste or odor characteristics.

Boltz's biomolecular models will be combined with dsm-firmenich's receptor data and sensory-science expertise to explore these interactions and support the design of new ingredients.

Exclusive AI Models for dsm-firmenich

The partnership is expected to produce a set of exclusive models incorporating dsm-firmenich's decades of knowledge in sensory sciences.

These models will be used for:

• Predicting molecular structures

• Designing new molecules

• Studying molecule-receptor interactions

• Exploring potential taste characteristics

• Exploring potential odor characteristics

• Identifying promising ingredient candidates

• Supporting faster ingredient discovery

The approach could allow researchers to evaluate molecular possibilities computationally before committing resources to laboratory synthesis and testing.

Scientists Will Access the Boltz Platform

Under the partnership, dsm-firmenich teams will have access to the Boltz platform, including the Boltz Lab interface and Boltz API.

The collaboration also includes an integration with large language model (LLM) agents, allowing scientists to orchestrate complex design and prediction workflows using natural-language instructions.

Embedded Boltz researchers will also work alongside dsm-firmenich teams. This arrangement is intended to combine AI development expertise with practical knowledge in chemistry, sensory science and ingredient discovery.

Accelerating Flavor and Fragrance Innovation

One of the major objectives of the partnership is to increase the speed at which new ingredients can be explored.

Traditional ingredient discovery can require extensive experimentation. Researchers need to identify promising molecular structures, synthesize compounds, evaluate their properties and conduct further testing.

AI can support this process by narrowing down large molecular spaces and identifying candidates that may warrant further investigation.

dsm-firmenich says its AI capabilities allow perfumers, flavorists and scientists to explore larger numbers of molecular possibilities and identify promising directions more quickly.

dsm-firmenich's Broader AI Strategy

The Boltz partnership is part of dsm-firmenich's broader strategy of using AI not only to improve operational efficiency but also to support scientific creativity and innovation.

The company has developed tools such as Molecule Fingerprinting, which uses AI to create digital fingerprints for molecules and predict characteristics related to scent, taste, stability and sustainability.

According to dsm-firmenich, these technologies are being used to help its scientists and creators explore chemical space and identify potential ingredients before they are produced in the laboratory.

The company refers to this broader approach as Discovery 2.0, combining AI, data science, chemistry and receptor biology in the ingredient-discovery process.

From AI Efficiency to AI-Driven Creativity

The partnership highlights an important shift in how AI is being applied in the flavor and fragrance industry.

AI is often used to automate repetitive tasks or analyze data. In ingredient discovery, however, the technology can also be used to generate new possibilities.

For perfumers and flavorists, this could mean exploring molecular structures that might not have been obvious through conventional approaches.

Human expertise remains important because AI-generated molecular candidates still need scientific evaluation, laboratory testing and sensory assessment. The partnership is therefore designed to combine computational capabilities with the creative and scientific expertise of dsm-firmenich teams.

Potential Impact on the Flavor and Fragrance Industry

AI-powered molecular discovery could have broader implications for the global flavor and fragrance industry.

Companies are continuously looking for new ingredients that can deliver specific sensory properties while meeting requirements related to performance, safety, availability and sustainability.

More efficient discovery tools could help researchers explore these requirements simultaneously and identify promising candidates earlier in the development process.

For dsm-firmenich, combining its proprietary sensory data with advanced biomolecular AI models could create a specialized technology platform focused on the unique challenges of taste and odor science.

Boltz Expands Its Biomolecular AI Applications

The partnership also represents an expansion of Boltz's biomolecular AI technology into the flavor and fragrance sector.

Boltz was founded by former MIT scientists and develops AI models for biomolecular structure prediction, molecular interaction modeling and generative design. Its technology has applications across pharmaceutical and biotechnology research.

Working with dsm-firmenich gives Boltz an opportunity to apply its technology to taste and odor receptors, which the company describes as some of the most interesting and understudied proteins in the human body.

What Comes Next?

The companies will work together to integrate Boltz's AI models into dsm-firmenich's ingredient-discovery workflows.

The focus will be on combining computational molecular design with dsm-firmenich's proprietary data, receptor biology expertise and sensory-science capabilities.

As the technology develops, the collaboration could help researchers identify and evaluate new molecules for flavor and fragrance applications with greater speed and precision.

Conclusion

The partnership between dsm-firmenich and Boltz represents a significant development in the use of AI for flavor and fragrance ingredient discovery.

By combining Boltz's frontier biomolecular AI models with dsm-firmenich's extensive sensory-science knowledge and taste and odor receptor data, the companies aim to create new tools for molecular design and ingredient discovery.

The collaboration will give dsm-firmenich scientists access to the Boltz platform, API and AI-agent capabilities while bringing Boltz researchers directly into the company's ingredient-discovery teams.

As AI continues to develop across the chemical, food, fragrance and life-science industries, partnerships such as this could reshape how new molecules are identified, designed and tested. For dsm-firmenich, the goal is to combine AI, scientific expertise and human creativity to accelerate the development of the next generation of flavor and fragrance ingredients.