Menu

LUNA AI

Scalable coprocessors
for decentralized compute

Learn More

Luna AI

The GPU for onchain execution ,

Luna AI allows computations to be delegated to offchain microservices with proofs and attestations verified onchain
Image 1

Verifiable offchain execution doesn't require replication of the entire computation and can thus provide results comparatively faster.

Image 2

Onchain verification of proofs or attestations is usually orders of magnitude cheaper than carrying out the entire computation onchain.

Image 3

Data and logic of computations carried out offchain are not visible onchain or to the host machine in the case of enclaves and cryptographic schemes.

MODULAR

We know one size doesn't fit all

Easily express your trust, cost and performance preferences

Branding

Ability to run trustless servers offchain enable transparent onchain execution abstracting away the underlying chain

Modular x Integrated

Easily plug-and-play coprocessors on any chain via a relay contract and offer users a uniquely integrated experience

Scalable

Auto-scaling and fault tolerance allow applications to seamlessly cater to increasing workloads without failing

EXPRESSIVE

Write code using familiar languages and frameworks. Execute verifiable programs on a serverless infrastructure that scales easily.

Fast and Secure Computation

See Our Featured Projects.

Quickly build, deploy, and scale apps with support for Python, C++, Go, and Solana. Keep data private using confidential computing.

code
Build

Fast development cycles Port existing Python, C++ or Go apps or use your favourite zkVM

Build

Fast development cycles Port existing Python, C++ or Go apps or use your favourite zkVM

Shutterbug
Easy maintenance

Seamlessly import Rust crates, ML or crypto libraries for better auditability

Easy maintenance

Seamlessly import Rust crates, ML or crypto libraries for better auditability

Explore
Fire-and-forget

Deploy your programs reliably with a single click and scale with demand

Fire-and-forget

Deploy your programs reliably with a single click and scale with demand

Minimalismo
Support private inputs

Keep models and prompts private via confidential computing on H100s

Support private inputs

Keep models and prompts private via confidential computing on H100s

Explore the ecosystem

Discover the vast variety of dApps that can leverage Luna AI coprocessors

  • Run risk assessment algorithms for lending or margin protocols, develop recommendation systems for DeSoc or develop autonomous worlds with dynamic environments

    Author image
    AI models Luna AI.