Models / Gemma 3 4B

Gemma 3 4B

edge
HuggingFace → Google → Find for my hardware

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 coding
Parameters
4.3B
Context
128k
License
other
Developer
Google
Origin
🇺🇸 USA
Released
Feb 2025

Scores

Coding
42
Reasoning
45
General
55

Score per dollar

1375 pts per $/M input

general_score (55) divided by cheapest input price ($0.04/M). Higher is better value. See live pricing.

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 3n E4B
4.0B edge 8GB 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

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
3.0GB 3.0GB min 3.5GB rec
Balanced - the usual local sweet spot
Q8_0
5.0GB 5.0GB min 5.5GB 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 2455.7t/s fast 1473.7t/s
NVIDIA DGX Station 748GB fast 1466.1t/s fast 879.8t/s
8x RTX 3090 rack (192GB) fast 1372.5t/s fast 823.7t/s
4x RTX 5090 (128GB) fast 1313.6t/s fast 788.3t/s
AMD Instinct MI300X (192GB) fast 975.8t/s fast 585.6t/s
4x RTX 4090 (96GB) fast 738.9t/s fast 443.4t/s
2x RTX 5090 (64GB) fast 656.8t/s fast 394.2t/s
2x RTX 3090 (48GB) fast 343.1t/s fast 205.9t/s
Single RTX 5090 (32GB) fast 328.4t/s fast 197.1t/s
RTX PRO 6000 Blackwell (96GB) fast 328.4t/s fast 197.1t/s
Mac Studio M4 Ultra 192GB fast 218.3t/s fast 131.0t/s
Mac Studio M4 Ultra 512GB fast 218.3t/s fast 131.0t/s
Single RTX 4090 (24GB) fast 184.7t/s fast 110.9t/s
MacBook Pro M5 Max 128GB fast 122.8t/s fast 73.7t/s
Single GTX 1080 Ti (11GB) fast 88.7t/s fast 53.2t/s
Dual EPYC 9004 + 768GB DDR5-4800 fast 84.5t/s fast 50.7t/s
DGX Spark 128GB unified fast 50.0t/s fast 30.0t/s
Ryzen AI Max+ 395 128GB fast 46.9t/s fast 28.2t/s
Jetson AGX Orin 64GB fast 37.5t/s fast 22.5t/s
Epyc + 512GB DDR4-3200 + 2x RTX 3090 fast 37.5t/s fast 22.5t/s
Epyc + 512GB DDR4-2400 + 2x RTX 3090 fast 28.2t/s ok 16.9t/s
NVIDIA Jetson Orin NX 16GB ok 18.8t/s ok 11.3t/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
3.0GB dl 3.0GB min 3.5GB rec
REC RAM vs largest quant
Q8_0 official -1% vs fp16
5.0GB dl 5.0GB min 5.5GB rec
REC RAM vs largest quant

Or run it in the cloud

Live per-provider pricing, throughput and uptime - refreshed about 19 hours ago via OpenRouter. Click a column to sort.

Provider Type Input $/M Output $/M Cache $/M Tok/s Latency Uptime Value
Parasail
API 0.04 0.05 0.020 - - 100.00% best uptime
DeepInfra
API 0.05 0.10 0.020 - - 100.00%

Default order: throughput among 95%+ uptime providers, then latency; subscriptions last. Sort by any column. Subscription rows show $/mo in the Value column - per-token columns are "-". Affiliate links are marked sponsored / nofollow. Confirm current pricing on the provider's site before committing.

Detailed API pricing page + JSON endpoint →

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...