AIMarkTechPostabout 1 hour ago
Needle 2: 45M-Parameter Model in 14MB Binary
9 min read
Cactus Compute released Needle 2, an open 45M-parameter tool-calling model shipping as a single 14MB binary. It runs a full session in about 28MB of RAM, with decode throughput of 500 tokens/sec on a Raspberry Pi 5.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
