§01·compatibility · /check
Gemma 4 26B MoE on RTX 3090
Yes — Gemma 4 26B MoE runs on the RTX 3090 (24 GB). Fastest community-measured result: 3625.6 prefill tokens/s, with 24 GB peak VRAM.
✓ runsmultimodalactive30 series24GB VRAM
model
- name
- Gemma 4 26B MoE
- slug
- gemma4-26b
- vertical
- multimodal
- status
- active
- repo
- huggingface.co ↗
§02·benchmarks
| Task | Quant | Speed | VRAM | Works | Confidence | Source | Verified |
|---|---|---|---|---|---|---|---|
| llm | Q4_K | 3625.6prefill tokens/s | 24GB | ✓ | hardware-corner.net· web | 2026-05-15 | |
| llm | Q4_K | 119.4tokens/s | 24GB | ✓ | hardware-corner.net· web | 2026-05-15 |
§03·common questions
Can you run Gemma 4 26B MoE on RTX 3090?
Yes — Gemma 4 26B MoE runs on the RTX 3090 (24 GB). Fastest community-measured result: 3625.6 prefill tokens/s, with 24 GB peak VRAM.
How much VRAM does Gemma 4 26B MoE need on RTX 3090?
Measured peak VRAM is 24 GB.
Which quantizations have been tested for Gemma 4 26B MoE on RTX 3090?
Q4_K — measured in community benchmarks.
How fast is Gemma 4 26B MoE on RTX 3090?
Up to 3625.6 prefill tokens/s (llm), the fastest community-measured result.
Are there step-by-step instructions for Gemma 4 26B MoE on RTX 3090?
Yes — a step-by-step recipe documents Gemma 4 26B MoE on the RTX 3090; see the recipes listed below.
§04·related recipes
- llmintermediate18GB+
Gemma 4 26B A4B-it on RTX 3090: Local Multimodal Chat via Q4_K_M GGUF + llama.cpp
- llmintermediate18GB+
Gemma 4 26B A4B-it on RTX 3090 Ti: Local Multimodal Chat via Q4_K_M GGUF + llama.cpp
- llmintermediate29GB+
Gemma 4 26B A4B-it on RTX 5090: Q8_0 Quality Tier via ggml-org GGUF + llama.cpp
- llmintermediate18GB+
Gemma 4 26B A4B-it on RTX 4090: Local Multimodal Chat via Q4_K_M GGUF + llama.cpp