Models / Gemma 3n E4B

Gemma 3n E4B

edge
HuggingFace → Google → Find for my hardware

Gemma 3n E4B: a raw 8B weight set run at an effective 4B footprint (~3GB in RAM) via Per-Layer Embeddings + MatFormer. The E2B model is nested inside it (Mix-n-Match can emit intermediate sizes). Multimodal - text, image, audio, video in; text out. 32K context. First sub-10B model to break 1300 on LMArena. Ollama: gemma3n:e4b-it-q4_K_M (7.5GB).

AI-generated content marks

The provider reports that this model does not add embedded watermarks or provenance metadata to generated output.

Provider transparency docs →

chat general vision
Parameters
4.0B
Context
32k
License
other
Developer
Google
Origin
🇺🇸 USA
Released
Jun 2025

Scores

Coding
46
Reasoning
48
General
58

Related models

Gemma 4 26B A4B
25.2B 3.8B active enthusiast 15GB min RAM $0.04/M in Apr 30, 2026
Gemma 4 31B
30.7B enthusiast 19GB min RAM $0.10/M in Apr 2, 2026
Gemma 3 270M
270M edge 1GB min RAM Aug 14, 2025
Gemma 3n E2B
2.0B edge 6GB min RAM Jun 26, 2025
Gemma 4 12B
12.0B consumer 7GB min RAM Jun 1, 2025
Gemma 3 12B
12.2B consumer 8GB min RAM $0.08/M in Feb 27, 2025
Gemma 3 1B
1.0B edge 1GB min RAM Feb 27, 2025
Gemma 3 27B
27.2B consumer 17GB min RAM $0.08/M in Feb 27, 2025
Gemma 3 4B
4.3B edge 3GB min RAM $0.04/M in Feb 27, 2025

Save your hardware and every model page answers the real question: will it run on your machine, and how fast?

Join free - save your rig →

Run it locally

Per-quant memory needs and a static "can you run it?" reference - no rig entry required

Q4_K_M
7.5GB 8.0GB min 10.0GB rec
Balanced - the usual local sweet spot
Q8_0
9.5GB 10.0GB min 12.0GB rec
Near-lossless

The reference hardware

Schematic of the 4x H100 80GB (320GB) reference rig - 320GB VRAM, 13400 GB/s aggregate bandwidth
Schematic of the NVIDIA DGX Station 748GB reference rig - 748GB unified memory, 8000 GB/s aggregate bandwidth
Schematic of the 8x RTX 3090 rack (192GB) reference rig - 192GB VRAM, 7489 GB/s aggregate bandwidth
Schematic of the 4x RTX 5090 (128GB) reference rig - 128GB VRAM, 7168 GB/s aggregate bandwidth
Schematic of the AMD Instinct MI300X (192GB) reference rig - 192GB VRAM, 5324 GB/s aggregate bandwidth
Schematic of the 4x RTX 4090 (96GB) reference rig - 96GB VRAM, 4032 GB/s aggregate bandwidth
Schematic of the 2x RTX 5090 (64GB) reference rig - 64GB VRAM, 3584 GB/s aggregate bandwidth
Schematic of the 2x RTX 3090 (48GB) reference rig - 48GB VRAM, 1872 GB/s aggregate bandwidth
Schematic of the Single RTX 5090 (32GB) reference rig - 32GB VRAM, 1792 GB/s aggregate bandwidth
Schematic of the RTX PRO 6000 Blackwell (96GB) reference rig - 96GB VRAM, 1792 GB/s aggregate bandwidth
Schematic of the Mac Studio M4 Ultra 192GB reference rig - 192GB unified memory, 1092 GB/s aggregate bandwidth
Schematic of the Mac Studio M4 Ultra 512GB reference rig - 512GB unified memory, 1092 GB/s aggregate bandwidth
Schematic of the Single RTX 4090 (24GB) reference rig - 24GB VRAM, 1008 GB/s aggregate bandwidth
Schematic of the MacBook Pro M5 Max 128GB reference rig - 128GB unified memory, 614 GB/s aggregate bandwidth
Schematic of the Single GTX 1080 Ti (11GB) reference rig - 11GB VRAM, 484 GB/s aggregate bandwidth
Schematic of the Dual EPYC 9004 + 768GB DDR5-4800 reference rig - 768GB unified memory, 460 GB/s aggregate bandwidth
Schematic of the DGX Spark 128GB unified reference rig - 128GB unified memory, 273 GB/s aggregate bandwidth
Schematic of the Ryzen AI Max+ 395 128GB reference rig - 128GB unified memory, 256 GB/s aggregate bandwidth
Schematic of the Jetson AGX Orin 64GB reference rig - 64GB unified memory, 204 GB/s aggregate bandwidth
Schematic of the Epyc + 512GB DDR4-3200 + 2x RTX 3090 reference rig - 560GB unified memory, 204 GB/s aggregate bandwidth
Schematic of the Epyc + 512GB DDR4-2400 + 2x RTX 3090 reference rig - 560GB unified memory, 153 GB/s aggregate bandwidth
Schematic of the NVIDIA Jetson Orin NX 16GB reference rig - 16GB unified memory, 102 GB/s aggregate bandwidth

22 reference configs, drawn in-house. Scroll for more.

Can you run it? - reference rigs

Rig Q4_K_M Q8_0
4x H100 80GB (320GB) fast 982.5t/s fast 775.7t/s
NVIDIA DGX Station 748GB fast 586.6t/s fast 463.1t/s
8x RTX 3090 rack (192GB) fast 549.2t/s fast 433.6t/s
4x RTX 5090 (128GB) fast 525.6t/s fast 415.0t/s
AMD Instinct MI300X (192GB) fast 390.4t/s fast 308.3t/s
4x RTX 4090 (96GB) fast 295.6t/s fast 233.4t/s
2x RTX 5090 (64GB) fast 262.8t/s fast 207.5t/s
2x RTX 3090 (48GB) fast 137.3t/s fast 108.4t/s
Single RTX 5090 (32GB) fast 131.4t/s fast 103.7t/s
RTX PRO 6000 Blackwell (96GB) fast 131.4t/s fast 103.7t/s
Mac Studio M4 Ultra 192GB fast 87.4t/s fast 69.0t/s
Mac Studio M4 Ultra 512GB fast 87.4t/s fast 69.0t/s
Single RTX 4090 (24GB) fast 73.9t/s fast 58.4t/s
MacBook Pro M5 Max 128GB fast 49.1t/s fast 38.8t/s
Single GTX 1080 Ti (11GB) fast 35.5t/s tight
Dual EPYC 9004 + 768GB DDR5-4800 fast 33.8t/s fast 26.7t/s
DGX Spark 128GB unified fast 20.0t/s ok 15.8t/s
Ryzen AI Max+ 395 128GB ok 18.8t/s ok 14.8t/s
Jetson AGX Orin 64GB ok 15.0t/s ok 11.9t/s
Epyc + 512GB DDR4-3200 + 2x RTX 3090 ok 15.0t/s ok 11.9t/s
Epyc + 512GB DDR4-2400 + 2x RTX 3090 ok 11.3t/s ok 8.9t/s
NVIDIA Jetson Orin NX 16GB slow 7.5t/s slow 5.9t/s

Fit tiers use the same will-it-run logic as the rig finder. For comfortable fits, the badge reflects decode speed: fast >=20 t/s, ok 8-20 t/s, slow <8 t/s. t/s is a bandwidth estimate, not a measured benchmark.

Download options

Q4_K_M official -5% vs fp16
7.5GB dl 8.0GB min 10.0GB rec
REC RAM vs largest quant
Q8_0 official -1% vs fp16
9.5GB dl 10.0GB min 12.0GB rec
REC RAM vs largest quant

Or run it in the cloud

No per-token API provider pricing tracked for Gemma 3n E4B yet. For flagship list prices, see the calculator.

See who runs Google in production →

PRICE HISTORY

Inference cost over time

Data accumulates from the first daily sync - longer ranges populate over time. Prices come from OpenRouter snapshots, not a historical API.

Loading price history...