Models / Gemma 4 E4B

Best home rig for Gemma 4 E4B

Quick answer

What is the fastest measured hardware for Gemma 4 E4B?
RTX 5070 Ti: 178.6 tok/s generation (llama.cpp b12128 (CUDA), Q4_K_M), from https://gearvn.com/blogs/danh-gia-tu-van/so-sanh-gpu-ai-local-b70-r9700-rtx-5070-ti.

Last updated: 11 Nov 2026

Legacy model2 measured resultsfastest plain decode: 178.6 tok/s

Older model. For current picks see current models.

Real, measured generation speeds for Gemma 4 E4B on home hardware, fastest first. Different quants/engines aren't directly comparable.

Build	Quant	Engine	tok/s	Prompt processing (tok/s)	Cost	Source
RTX 5070 Ti 16 GB (Core Ultra 9 285K, llama.cpp b12128 CUDA): gpt-oss 20B at 236.7 tok/s, Gemma 4 E4B Q4_K_M 178.6, Qwen3.5 9B Q6_K 107.5	Q4_K_M	llama.cpp b12128 (CUDA)	178.6	10,724 @4,096 · est. TTFT @8k ≈0.76 s*		gearvn.com ↗
Intel Arc Pro B70 32 GB (Core i7-12700KF, llama.cpp b12128 SYCL): gpt-oss 20B MXFP4 at 134.2 tok/s, Gemma 4 E4B Q4_K_M 97.3, Qwen3.5 9B Q6_K 66.5	Q4_K_M	llama.cpp b12128 (SYCL)	97.3	4,610 @4,096 · est. TTFT @8k ≈1.8 s*		gearvn.com ↗

Prompt processing = prefill speed reported by the source, @ prompt length. * est. TTFT @8k is calculated from measured prefill, not measured. Why it matters.

Rent before you buy: try this model on a cloud GPU for a few dollars first: Runpod (referral link; sign up with Google to get the new-user bonus)

FAQ

What is the fastest measured hardware for Gemma 4 E4B?

RTX 5070 Ti: 178.6 tok/s generation (llama.cpp b12128 (CUDA), Q4_K_M), from https://gearvn.com/blogs/danh-gia-tu-van/so-sanh-gpu-ai-local-b70-r9700-rtx-5070-ti.