PrismML is an artificial intelligence lab building minuscule models with exemplary power. With lower overhead, negligible latency, and affordable costs, our models are designed for the next wave of AI-powered devices.
Bonsai is our flagship model. Engineered with robotics, real-time agents, and edge computing in mind, Bonsai competes with LLaMA and gpt-oss with a 10x smaller footprint. Requiring only 1.1GB in RAM, Bonsai can comfortably run on smartphones, IoT devices, and other common appliances.
1-bit model fits on an iPhone 12, Raspberry Pi, or cheap Android
Centering AI research on efficiency
PrismML was founded on a mission of shrinking models without compromising on performance. Today’s LLMs helped jettison innovation into an AI-forward age, but are computationally draining on today’s resources and hardware. PrismML’s role is to bridge modern-day intelligence into an economical form-factor.
Join us to turn mathematically rigorous research into models that run anywhere. If you’re excited to push the frontier of efficient intelligence and ship it into the real world, we’d love to work with you.
Join us to turn mathematically rigorous research into models that run anywhere. If you’re excited to push the frontier of efficient intelligence and ship it into the real world, we’d love to work with you.
Thanks, your application has been received!
Oops! Something went wrong while submitting the form.
Contact us
Have a question, partnership idea, or a project that needs efficient intelligence? Reach out—we’d love to hear from you.
Thanks, your message has been received!
Oops! Something went wrong while submitting the form.