Liquid AI
Device-native foundation models for the physical world.
Liquid AI is an efficiency-first foundation model company that develops device-native artificial intelligence. Spun out of MIT CSAIL, the company builds Liquid Foundation Models (LFMs) designed to run on processors outside of traditional data centers, addressing the latency, privacy, and hardware constraints of the physical world. Its technology is built on the principle of co-designing model architectures with target hardware in the loop to optimize for memory and power efficiency.
The company provides a range of multimodal models including text, vision, audio, and specialized decision models. These models are engineered to be "lifelong learners" that can improve after deployment. Liquid AI's offerings serve diverse industries such as automotive, financial services, and biotechnology, enabling on-device applications like in-car assistants and private pharmaceutical research tools.
Based in Cambridge, Massachusetts, Liquid AI offers its models through an API for cloud-based access and as open-weight models for local deployment. The company partners with major silicon providers and global technology groups to scale its efficient AI solutions across consumer and enterprise devices.
- Headquarters
- Cambridge, Massachusetts
- Website: liquid.ai
Industries: Artificial Intelligence, Technology
Categories: AI Infrastructure, Generative AI
AntSpace put this profile together from public information. It may be incomplete; once Liquid AI claims it, its admins can edit everything here.
Our vision
Liquid AI's vision is that intelligence should be ubiquitous, co-designed with hardware, and never static, evolving into lifelong learners that adapt to each user.
Our culture
Liquid AI operates with a team that values merit and evidence over the source of an idea, maintaining a high bar for impact and results. The company encourages working from first principles and staying comfortable with ambiguity, granting colleagues the autonomy to drive their own work. Trust is a central pillar, with experts expected to execute autonomously while being open to early feedback and peer scrutiny. The culture prioritizes documented, traceable, and legible decision-making over black-box shortcuts. Liquid AI describes itself as a product-driven, customer-first business that respects employees' time and treats them as equals, aligning company goals with personal and professional growth.
Core values
- Product-driven
- Customer-first
Specialties
- Liquid Foundation Models
- Hardware-in-the-loop co-design
- Decision models
- Edge AI platform
- Scientific foundation models
Markets served
- Artificial Intelligence
- Automotive
- Financial Services
- Biotechnology
- Consumer Electronics
- Defense
Key facts
- Legal name
- Liquid AI, Inc.
Latest from Liquid AI
- Introducing LFM3: Every Modality for Specialized Intelligence (Product launch · October 27, 2026)
Liquid AI introduced LFM3, a new family of foundation models designed for specialized intelligence across multiple modalities including text, vision, and audio, optimized for on-device performance. - d1-ultra: Introducing the World’s Best Decision Model (Product launch · October 15, 2026)
The company released d1-ultra, a high-performance decision model that provides calibrated probabilities for classification and scoring tasks across text and vision inputs. - Liquid AI and Mercedes-Benz partner to scale embedded in-car intelligence (Partnership · April 23, 2026)
Liquid AI and Mercedes-Benz entered a partnership to integrate embedded intelligence into vehicles, enabling natural voice interactions and cabin vision without cloud dependencies.
Products by Liquid AI
- liquid-finetune (Developer tool)
Official fine-tuning framework for Liquid Foundation Models.
Services by Liquid AI
- Automated Foundation Model Design (AFMD) (Managed service)
The right architecture, co‑designed on silicon.
AI models by Liquid AI
- LFM2-2.6B-MMAI (Science / biology model)
Lightweight Scientific Foundation Model for Drug Discovery
Clients and partners
4 clients and partners
Case studies
2 case studies
Documents
2 documents
5 more products, services and AI on AntSpace. Sign up free to see the full catalog.
Leadership, clients, case studies, documents and every location on AntSpace. Sign up free to see them.