168 reports. These are individual setups, not a controlled benchmark. Gemma 4 speeds by hardware →
NVIDIA RTX 4070 · Ollama
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports about 55 t/s on an RTX 4070 12GB. The figure is estimated from compute-market tiers rather than a measured run.
NVIDIA RTX 3060 12GB · Ollama
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports E2B at about 60 t/s on an RTX 3060 12GB. The figure is estimated from compute-market tiers rather than measured.
NVIDIA RTX 3060 12GB · Ollama
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports about 45 t/s on an RTX 3060 12GB. The E4B model fits easily.
NVIDIA RTX 5090 · 50,000 ctx
User reports nvidia/Gemma-4-26B-A4B-NVFP4 running on an RTX 5090 with 80% of the 32 GB allocated, reaching around 50k context. The model is 18.8 GB. Benchmarks compare baseline against NVFP4: GPQA Diamond 80.30% vs 79.90%, AIME 2025 88.95% vs 90.00%, MMLU Pro 85.00% vs 84.80%, LiveCodeBench pass@1 80.50% vs 79.80%, IFBench 77.77% vs 78.1%, and IFEval 96.60% vs 96.40%.
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports about 150 t/s generation. The model is described as a star performer.
NVIDIA RTX A6000 48GB · llama.cpp
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports bf16 with no quantization at 10.25 GB VRAM and 61 ms TTFT. Source is dev.to Gaurav Vij.
M1 8GB · Ollama
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is shown only when the source reports it; generation speed cannot tell us time to first token. Check the full setup before comparing.
User reports Gemma 4 at 15-20 t/s, a range the user calls usable for simple tasks.
m5 pro (18 core cpu, 20 core gpu) · Ollama
Reported by the source; GPU count, offloading and concurrent requests can change these figures. Prompt-processing speed is input throughput, not time to first token. Check the full setup before comparing.
User reports 204.07 t/s prompt eval and 68.76 t/s generation on an M5 Pro with 48 GB RAM.