Model record
Gemma 4 E4B
Google DeepMind · Apache 2.0 · verified
What this model is
E4B means 4.5B effective parameters. The full checkpoint is about 8B parameters because the Per-Layer Embedding tables are stored even though they are lookup-only.
The architecture is dense. Every parameter is part of the active path, so memory and speed scale with the full model rather than a smaller expert subset.
Memory and quantization
Decimal GB. Q4_0 uses Google's QAT GGUF; other sizes use the published Unsloth GGUF. All totals include the multimodal projector.
| Quantization | Published files | 8K estimate | Smallest comfortable GPU | Smallest comfortable Mac | Context at smallest GPU / Mac | Speed estimate | Run commands |
|---|---|---|---|---|---|---|---|
| Q4_K_M | 5.97 GB | 7.74 GB | NVIDIA GeForce RTX 3060 12GB · 12 GB VRAM | Apple Mac mini M4 16GB · 16 GB unified memory | 131,072 / 131,072 tokens | Estimate · Fast relative class | ollama run gemma4:e4b-it-q4_K_M lms get "Gemma 4 E4B"@q4_k_m |
| Q4_0 | 6.15 GB | 7.92 GB | NVIDIA GeForce RTX 3060 12GB · 12 GB VRAM | Apple Mac mini M4 16GB · 16 GB unified memory | 131,072 / 131,072 tokens | Estimate · Fast relative class | ollama run gemma4:e4b-it-qat lms get "Gemma 4 E4B"@q4_0 |
| Q8_0 | 9.19 GB | 10.96 GB | NVIDIA GeForce RTX 4060 Ti 16GB · 16 GB VRAM | Apple Mac mini M4 16GB · 16 GB unified memory | 131,072 / 131,072 tokens | Estimate · Fast relative class | ollama run gemma4:e4b-it-q8_0 lms get "Gemma 4 E4B"@q8_0 |
Estimate formula: published weight files + 0.2716 GB per 8K tokens × context / 8K + 1.5 GB runtime overhead. Sizes are decimal GB, not GiB. At 8K context, the smallest listed build is about 7.7 GB before OS and application reserve.
Conservative FP16 estimate from the public config. Seven of 42 layers use global attention and local layers use a 512-token sliding window.
Ollama tags were checked against the linked Ollama library pages on 2026-11-02. LM Studio's lms get suffix selects a file when the search results expose that exact quantization; confirm the local key with lms ls.
KV method source: https://huggingface.co/google/gemma-4-E4B-it/raw/main/config.json
Speed heuristic
Not benchmarked here. The calculator uses the 4.5B effective parameter count only as a relative speed heuristic.
Active path: 4.5B parameters. This is a relative speed signal only; it is not a tokens/sec benchmark.
Runtimes
- Ollama
- llama.cpp
- MLX
ollama run gemma4:e4blms get "Gemma 4 E4B"@q4_k_m
lms load <model-key-from-lms-ls> --gpu max --context-length 8192License restrictions
Apache 2.0; the separate Gemma Prohibited Use Policy and intended-use statement still apply.
This is a plain-language summary, not legal advice. Read the linked license before commercial deployment or redistribution.
Known issues
- The download is closer to an 8B model than a 4.5B model because the embedding tables are still stored.
- Vision and audio support require the multimodal projector files.
If it does not fit, try these smaller records
These alternatives have a lower recommended memory target in the current catalog. Open each page to compare the quantization-specific fit before downloading.
Phi-4 Mini
Dense · starting at 5 GB
Llama 3.2 3B
Dense · starting at 4 GB
FAQ
Model-specific questions
What is the smallest local file for Gemma 4 E4B?
The smallest published weight package in this record is Q4_K_M at 5.97 GB. The 8K runtime estimate adds 0.27 GB of KV cache and 1.5 GB overhead, for about 7.7 GB before operating-system reserve.
How much memory do the published quantizations need at 8K?
Q4_K_M about 7.7 GB at 8K; Q4_0 about 7.9 GB at 8K; Q8_0 about 11.0 GB at 8K. These are calculated estimates using the published weight sizes and the model's KV-cache assumption, not live memory measurements.
Which GPU and Mac configurations are comfortable for Gemma 4 E4B?
Against the current catalog, the smallest configurations that clear the 20% headroom target are: Q4_K_M: NVIDIA GeForce RTX 3060 12GB; Q4_0: NVIDIA GeForce RTX 3060 12GB; Q8_0: NVIDIA GeForce RTX 4060 Ti 16GB. A larger configuration may still be preferable for longer context or faster memory bandwidth.
What is the right Ollama or LM Studio command for Gemma 4 E4B?
Use the command attached to the specific quantization in the table. Ollama tags are taken from the checked Ollama library pages. For LM Studio, use lms get with the displayed quantization suffix, then load the resulting local key from lms ls with --gpu max and the desired context length.
What should I run instead if Gemma 4 E4B does not fit?
Smaller catalog alternatives include Phi-4 Mini, Llama 3.2 3B. Compare their quantization tables and minimum memory before choosing; a smaller parameter count does not always mean a proportionally smaller download.
Sources
- https://huggingface.co/google/gemma-4-E4B-it
- https://huggingface.co/google/gemma-4-E4B-it-qat-q4_0-gguf
- https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF
- https://ollama.com/library/gemma4
- https://huggingface.co/google/gemma-4-E4B-it/raw/main/config.json
This page is an independent compatibility guide, not official documentation. Verify the primary source before downloading or deploying a model.