Cactus Engine
The runtime for consumer hardware

The source-available on-device runtime behind everything we ship.

Inspect the code that runs on your users' devices, available under the Cactus license.

$git clone git@github.com/cactus-compute/cactus

$source ./setup

$cactus build

$cactus run LiquidAI/LFM2-2.6B

The hybrid router isn't locked to our engine. Keep the runtime you already ship — Cactus routes around it.

Quantized models with hardware-specific acceleration. Tuned for battery-efficient inference.

iOS, Android, macOS, and wearables from a single SDK. Write once, deploy anywhere.

Built by a team from

Y CombinatorUniversity of OxfordDeepRenderSalesforceGoogleAWSWashington PostMIT

Ready to get started?

Add hybrid inference to your app in minutes. Free to start, scales with you.