WTF Is Going On(Weights, Tools & Frameworks)

한국어
google/embeddinggemma-2
feature-extraction apache-2.0 GGUFMLXNVFP4

What this is

744M parameters. Turns input into vectors for search and retrieval. Not something you chat with.

Packaged as GGUF, MLX, NVFP4, the usual download about 0.3 GB. llama.cpp, Ollama and LM Studio read GGUF.

3 independent publishers have repackaged it since it appeared.

apache-2.0 — commercial use allowed.

Release chainonly the steps this archive can prove

  1. weights posted2026-09-14
  2. runnable3 publishers

Measured herefrom our own daily capture — upstream publishes today only

Who republished itderivatives we have seen, by download size

← the board