DeepYard
N

Needle

Lightweight tool-calling model optimized for edge deployment and WebAssembly runtime

Open SourceFree

About

Needle is a specialized function-calling model from Cactus Compute designed for on-device agent execution. Built with edge deployment in mind, it supports WebAssembly runtime and advanced quantization techniques, enabling AI agents to run directly on user devices without cloud dependencies. Ideal for developers building privacy-focused, low-latency agent applications that need to operate offline or in resource-constrained environments.

Details

Type
Integrations
Language

Tags

tool-useself-hostedopen-sourceautonomouscli