self-hosted/ai
§01·model · /models

Gemma 4 E4B-IT

multimodalactiveApache-2.0

Multimodal instruction-tuned model by Google (Gemma 4 E4B-IT) — an efficiency-optimised ~4B-effective variant handling text and vision, Apache-2.0. It carries a recipe on every one of the 27 catalogued GPUs — NVIDIA, AMD and Apple alike, from an 8 GB floor upward, one of only three models here with that full sweep. Measured at 45.0 tok/s on a 12 GB RTX 3060. If you want one model that will run on whatever card is actually in the machine, this is a safe pick.

§02·same family
3 other models · by name

Other models grouped with Gemma 4 E4B-IT. Sizes and modalities can differ, and nothing here says whether one fits your GPU — open a model for its own compatibility table.

§03·GPUs that run this model
27 total

benchmarked·~ runs via recipe (not benchmarked)· untested·doesn't fit