Google
ChatCodingReasoning8B
Chat

Gemma 4 E4B

Gemma · Apache 2.0

Gemma 4's mid-range edge model. 8B total parameters with 4.5B effective. Full multimodal: text, image, audio, and video. 52% LiveCodeBench v6 and 42.5% AIME 2026 put its reasoning and coding above most models at this size. Fits comfortably on 8 GB GPUs at Q4_K_M. Apache 2.0 licensed.

Parameters
8B
Architecture
Dense
Context
128,000 tokens
Released
2026-04-02
Engines
llama.cpp, ollama, vLLM, TGI
Builder Tools
Aider, Claude Code, Continue, Cursor, LM Studio, Ollama, Open WebUI, Windsurf

Parameters

8B

VRAM

7 GB

Context

125K

Formats

4

GPUs

43

Gemma 4 E4B (8B) requires 7 GB VRAM at recommended quality (Q6_K). At efficient quality (Q4_K_M), it fits in 6 GB VRAM, making it compatible with the NVIDIA RTX 4060 Laptop (40-60W). On NVIDIA Grace Blackwell Ultra GB300, expect approximately 500 tok/s at Q8_0. For the best experience, Starter AI Desktop ($582) is recommended.

Source: OwnRig methodology

VRAM (Recommended)

7 GB

Quantization

Q6_K

File Size

6.33 GB

Max Context

125K tokens

Primary Use

Chat

Memory

VRAM Requirements

QualityQuantizationVRAMFile Size
fullQ8_09 GB8.03 GB
recommendedQ6_K7 GB6.33 GB
recommendedQ5_K_M6.5 GB5.82 GB
efficientQ4_K_M6 GB5.41 GB
Scaling

Context Length Impact

KV cache VRAM at Q6_K quality. Longer context = more memory.

ContextKV CacheTotal VRAM
2K102 MB7.1 GB
4K205 MB7.2 GB
8K307 MB7.3 GB
16K717 MB7.7 GB
32K1.3 GB8.3 GB
64K2.7 GB9.7 GB

Compatible GPUs

43 devices
NVIDIA Grace Blackwell Ultra GB300Q8_0500 tok/sExcellent
Apple M4 Ultra (192GB)Q8_076 tok/sExcellent
AMD Radeon Pro W7900Q8_080 tok/sExcellent
NVIDIA GeForce RTX 3080 10GBQ8_070 tok/sExcellent
NVIDIA GeForce RTX 3090Q8_087 tok/sExcellent
NVIDIA GeForce RTX 4070 Ti SuperQ8_062 tok/sExcellent
NVIDIA GeForce RTX 4080 SuperQ8_068 tok/sExcellent
NVIDIA GeForce RTX 4090Q8_094 tok/sExcellent
NVIDIA GeForce RTX 5080Q8_089 tok/sExcellent
NVIDIA GeForce RTX 5090Q8_0167 tok/sExcellent
NVIDIA RTX PRO 6000 BlackwellQ8_0168 tok/sExcellent
NVIDIA RTX PRO 6000 Blackwell Max-QQ8_0168 tok/sExcellent
AMD Radeon RX 7900 XTXQ8_089 tok/sExcellent
Apple M4 Max (128GB Unified)Q8_050 tok/sGood
Apple M4 Max (36GB Unified)Q8_050 tok/sGood
Apple M4 Max (64GB Unified)Q8_050 tok/sGood
NVIDIA GeForce RTX 3060 12GBQ8_033 tok/sGood
NVIDIA GeForce RTX 4060 8GBQ6_K32 tok/sGood
NVIDIA RTX 4060 Laptop (40-60W)Q6_K30 tok/sGood
NVIDIA RTX 4070 Laptop (80-115W)Q6_K30 tok/sGood
NVIDIA GeForce RTX 4070 SuperQ8_047 tok/sGood
NVIDIA GeForce RTX 4070 Ti 12GBQ8_047 tok/sGood
NVIDIA RTX 4080 Laptop (120-150W)Q8_035 tok/sGood
NVIDIA RTX 4090 Laptop (150-175W)Q8_047 tok/sGood
AMD Radeon RX 7600Q6_K34 tok/sGood
AMD Radeon RX 9070Q8_058 tok/sGood
Apple M1 Pro (16GB Unified)Q8_014 tok/sGood
Apple M2 Pro (16GB Unified)Q8_016 tok/sGood
NVIDIA GeForce RTX 5060 8GBQ6_K37 tok/sGood
Apple M4 Pro (24GB Unified)Q8_025 tok/sAcceptable
Apple M4 Pro (48GB)Q8_025 tok/sAcceptable
NVIDIA GeForce RTX 4060 Ti 16GBQ8_026 tok/sAcceptable
AMD Radeon RX 9060 XT 16GBQ8_029 tok/sAcceptable
NVIDIA GeForce RTX 5060 Ti 16GBQ8_029 tok/sAcceptable
Apple M3 Pro (18GB Unified)Q8_014 tok/sMarginal
Apple M4 (16GB Unified)Q8_011 tok/sMarginal
Apple M1 (16GB Unified)Q8_03 tok/sMarginal
Apple M2 (16GB Unified)Q8_05 tok/sMarginal
Apple M3 (16GB Unified)Q8_06 tok/sMarginal
Apple M1 (8GB Unified)Q8_0Not viable
Apple M2 (8GB Unified)Q8_0Not viable
Apple M3 (8GB Unified)Q8_0Not viable
AMD Radeon RX 9060 XT 8GBQ8_0Not viable

Showing 43 of 43 entries

Builder Context

Gemma 4 E4B is commonly used with Aider, Claude Code, Continue, Cursor, LM Studio, Ollama, Open WebUI, Windsurf. For an AI coding workflow, pair it with an embedding model like nomic-embed-text for local RAG.

FAQ

Frequently Asked Questions

How much VRAM does Gemma 4 E4B need?
Gemma 4 E4B requires 7 GB VRAM at recommended quality (Q6_K). At lower quality settings, it can fit in as little as 6 GB.
What is the best GPU for Gemma 4 E4B?
The NVIDIA Grace Blackwell Ultra GB300 delivers the best performance for Gemma 4 E4B, achieving 500 tok/s at Q8_0 with an excellent rating.
Can I run Gemma 4 E4B on an RTX 4060 Ti?
Yes. On the NVIDIA GeForce RTX 4060 Ti 16GB, Gemma 4 E4B runs at 26 tok/s (Q8_0, acceptable).
What quantization should I use for Gemma 4 E4B?
For the best quality, use Q6_K (7 GB VRAM). If your GPU has limited VRAM, Q4_K_M (6 GB) is the most efficient option with acceptable quality.
Is Gemma 4 E4B good for coding?
Yes. Gemma 4 E4B is used with Aider, Claude Code, Continue, Cursor, LM Studio, Ollama, Open WebUI, Windsurf for local AI coding. For the best coding experience, pair it with an embedding model for local RAG.

Related Guides

All models

Data confidence: estimated. Source

VRAM requirements are calculated from model parameters and may vary by inference engine, context length, and batch size. Performance estimates are based on community benchmarks and should be verified for your specific configuration.Gemma is a trademark of its respective owner. OwnRig is not affiliated with or endorsed by the model creator.