back to local model checker

Gemma 4 E2B

Vision

Google · 5B · Dense

Gemma 4 efficient base model (official)

Gemma
Downloads

0

Likes

0

Released

2026-04

Context

128K tokens

Use cases

vision

Your hardware

16 GB Reference Desktop

Quant
Q4_K_M
Memory
3.1 GB
Speed
~65 tok/s
Fit
Runs great

Quantization options

What is quantization?
QuantBitsVRAMDiskQualityStatus
Q2_K22.1 GB1.5 GBlowRuns great · ~96 t/s
Q3_K_M32.7 GB2.1 GBmoderateRuns great · ~75 t/s
Q4_K_M43.1 GB2.4 GBgoodRuns great · ~65 t/s
Q5_K_M53.7 GB3.1 GBgoodRuns great · ~54 t/s
Q6_K64.3 GB3.7 GBexcellentRuns well · ~47 t/s
Q8_085.6 GB4.9 GBexcellentRuns well · ~36 t/s
F161610.7 GB9.8 GBlosslessTight fit · ~19 t/s

About this model

Gemma 4 E2B (Gemma 4) is a 5B dense transformer from Google in the Gemma family, designed primarily for vision workflows. Supports up to 128K context tokens; start at 4K–8K with Q8_0 KV cache on tight GPUs.

Hardware planning baselines start at 2.8 GB minimum memory and 4.7 GB recommended memory. At Q4_K_M, weights plus baseline runtime overhead use roughly 3.1 GB of RAM/VRAM and 2.4 GB of disk space. As a 5B dense checkpoint, every parameter participates in each token forward pass, meaning decoding speed scales linearly with GPU memory bandwidth and drops sharply if layers spill over PCIe into host DDR5 RAM.

Deploy Gemma 4 E2B locally via llama.cpp with Flash Attention enabled. Under Gemma Terms of Use, review official weight cards before commercial redistribution.

Highlights

  • Local Fit Baseline2.8 GB minimum and 4.7 GB recommended RAM/VRAM.
  • Primary QuantQ4_K_M targets 3.1 GB memory and 2.4 GB disk.
  • Context & Capabilities128K max context · Dense.

System requirements by device

Memory planning targets 2.8 GB minimum RAM/VRAM and 4.7 GB recommended memory. The baseline quant is Q4_K_M, estimated at 3.1 GB memory and 2.4 GB disk. Actual runtime usage shifts with context length, KV cache precision, and OS background overhead.

DeviceNotes
Mac (Apple silicon)Use unified memory headroom (~75% physical RAM max) as the primary fit signal. Start with Q4_K_M and close heavy browser tabs first.
Windows / Linux GPUKeep weights plus KV cache inside dedicated VRAM so Windows WDDM does not spill layers over PCIe into Shared GPU Memory.
AndroidFeasible on 8–16 GB phones via AI Edge Gallery or PocketPal when keeping context short (2K–4K) and closing background apps.
iPhone / iPadTarget 8 GB+ Pro iPhones or M-series iPads with Q4_K_M and 2K–4K context so allocations stay below iOS Jetsam ceilings.

Start with a short session and leave memory for the operating system. A file fitting on disk does not guarantee that it will load in memory.

Estimated memory by quantization

These are rough estimates, not tested minimum requirements. Check the actual file and runtime documentation; context, offloading and operating-system overhead can change the memory needed.

Quant TierEstimated memoryDisk SizeQuality
Q2_K2.1 GB1.5 GBlow
Q3_K_M2.7 GB2.1 GBmoderate
Q4_K_M3.1 GB2.4 GBgood
Q5_K_M3.7 GB3.1 GBgood
Q6_K4.3 GB3.7 GBexcellent
Q8_05.6 GB4.9 GBexcellent
F1610.7 GB9.8 GBlossless

How to run Gemma 4 E2B on your device

Start with a short prompt and the smallest quant that still gives the quality you need. A clean first run is more useful than loading a larger file with no memory headroom.

Mac

Run Gemma 4 E2B on Mac

  1. Install LM Studio or llama.cpp on macOS Apple Silicon.
  2. Pull Q4_K_M (3.1 GB) first and confirm it stays below ~75% of your Mac's unified memory.
  3. Enable Flash Attention and cap initial context at 8K until memory pressure stays below 4.7 GB.
  4. Consult the Mac guide for Activity Monitor tuning and Metal offload tips.
Device how-to guide
Windows

Run Gemma 4 E2B on Windows

  1. Install LM Studio for Windows for visual VRAM inspection, or Ollama when you want a localhost API endpoint.
  2. Load Q4_K_M (~3.1 GB) and verify Dedicated GPU Memory stays within VRAM without spilling into Shared GPU Memory.
  3. Reserve at least 2.8 GB memory minimum and target 4.7 GB for comfortable context growth.
  4. Enable Q8_0 KV cache or drop one quant tier if generation slows down over PCIe.
Device how-to guide
Android

Run Gemma 4 E2B on Android

  1. Use AI Edge Gallery or PocketPal AI on an Android phone with at least 8–12 GB RAM.
  2. Select Q4_K_M or a smaller mobile-sized quant (~3.1 GB) and keep context at 2K–4K.
  3. Close background camera and browser apps before loading so Android LMK does not kill the process.
  4. Consult the Android guide for SoC thermal and NPU/GPU delegate tips.
Device how-to guide
iPhone

Run Gemma 4 E2B on iPhone or iPad

  1. Install PocketPal or a Metal-enabled llama.cpp iOS client on an 8 GB+ iPhone or M-series iPad.
  2. Import Q4_K_M via Files and keep initial context at 2048 tokens to stay beneath iOS Jetsam limits.
  3. Test one short text prompt before running multi-turn chats.
  4. Move to a Mac if iOS terminates the app due to memory pressure.
Device how-to guide

Gemma 4 E2B vs alternatives

ModelParamsContextMin RAMLicense
Gemma 4 E2B5B128K2.8 GBGemma
Gemma 4 E2B IT5B128K2.8 GBApache 2.0
Gemma 3 4B4B128K2.2 GBGemma
Gemma 2 2B2B8K1.1 GBGemma
Gemma 4 E4B IT8B128K4.5 GBApache 2.0
Gemma 4 E4B8B128K4.5 GBGemma

Use this table to choose the smallest nearby model that still matches the job. A related model with lower RAM can be the better local choice when the current page fits only at the edge of available memory.

Frequently asked questions

What is Gemma 4 E2B?
Gemma 4 E2B is a 5B dense transformer from Google, supporting up to 128K tokens with 2.8 GB minimum RAM and 4.7 GB recommended RAM.
How do I download Gemma 4 on Android?
Start with an Android app that can import GGUF files, then test Gemma 4 E2B with Q4_K_M before trying larger quants. Use the hardware table on this page to confirm memory headroom first.
How much VRAM or RAM does Gemma 4 E2B require?
At Q4_K_M, Gemma 4 E2B occupies roughly 3.1 GB of active memory and 2.4 GB on disk. Add 2–6 GB of spare headroom for KV cache buffers, runtime scratch space, and OS apps (4.7 GB recommended).
What is the best way to run Gemma 4 E2B locally?
Run Gemma 4 E2B in llama.cpp starting with Q4_K_M. Keep initial context at 4K–8K tokens and confirm 100% GPU/Metal offload before scaling context length.
Can Gemma 4 E2B run on Android or iPhone?
Yes, on flagship 8–16 GB Android phones (via AI Edge Gallery or PocketPal) or 8 GB+ iPhones when capped to Q4_K_M and short 2K–4K context windows.
How does Gemma 4 E2B compare to Gemma 4 E2B IT?
Gemma 4 E2B IT (5B, 2.8 GB min RAM) sits closest in size. Compare active parameters, context ceiling (128K vs 128K), and license terms (Gemma) before switching.