Meet Needle 2: An Open 45M-Parameter Tool-Calling Model That Ships as a 14MB Binary and Runs a Full Session in 28MB of RAM

Cactus Compute has released Needle 2, an open-source 45M-parameter model designed for tool calling and structured data extraction. The model is notable for its lightweight design, as it functions as a 14MB binary and requires only 28MB of RAM to operate. Because it runs on devices lacking specialized GPU or NPU hardware, it offers a practical solution for deploying AI capabilities on constrained systems.