self-hosted/ai
§01·model · /models

Gemma 4 26B MoE

multimodalactiveApache-2.0

26B Mixture-of-Experts LLM by Google (Gemma 4, A4B - ~4B active per token), instruction-tuned. Apache-2.0.

§02·GPUs that run this model
4 total
GPUVRAMSeriesBest speedMin VRAMWorksBenchmarksRecipe
RTX 509032GB50180.3tokens/s2recipecheck ↗
RTX 309024GB30119.4tokens/s24GB2recipecheck ↗
RTX 3090 Ti24GB30~0recipecheck ↗
RTX 409024GB40~0recipecheck ↗

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