Gemma 4 E4B
Gemma 4 E4B is a small Google model for devices with little memory. It accepts text, images and audio and runs on a laptop with 16 GB, without a graphics card or internet connection. It is too small for long documents and demanding analysis, but often enough for short everyday tasks.
Profile
- Maker
- Google · United States
- Licence
- Free for commercial use
- Runs on
- Laptop with 16 GB
- Memory at 4 bits
- 6 GB
- Memory at 8 bits
- 8 GB
- Parameters
- 8bn
- Context
- 33,000 tokens
- Inputs
- text, image, audio
- Released
- April 2026
Scores
- EuroEval
- German: rank 57 of 140
Scores: LMArena, Epoch AI, EuroEval, Hugging Face, models.dev; as of 19 October 2026. Sources and method →
Use cases
Where the model makes sense.
Meeting notes in field sales
- Situation
- Field staff dictate notes after customer visits but often have no stable connection on the road.
- What it looks like
- The model runs on the laptop, takes the voice recording and turns it into a short summary with next steps.
- What to watch
- Check technical terms and names. For long recordings, a dedicated speech recognition model is more accurate.
Writing help where data is sensitive
- Situation
- At a practice reception or in HR, texts should be written faster without data leaving the device.
- What it looks like
- The model rephrases short emails, shortens texts or makes them friendlier, directly on the workstation.
- What to watch
- Small models make more grammar and factual errors. Every text is read before sending.
Photos of type plates and forms
- Situation
- Technicians have to copy serial numbers and device data from type plates.
- What it looks like
- A photo is enough; the model returns the details as text that goes into the job.
- What to watch
- Always double-check serial numbers; a swapped digit otherwise shows up late.
AI training without cloud accounts
- Situation
- Under the EU AI Act staff need AI literacy, and exercises should not use private cloud accounts.
- What it looks like
- Each participant works with the model on their own laptop and learns prompts, limits and typical errors.
- What to watch
- A small model shows weaknesses more clearly than large cloud models, which is rather an advantage for training.
The use cases were written with AI assistance (Claude by Anthropic) and reviewed by vona. Scores come unchanged from the sources listed.
When another model fits better
For long documents (the context window is small at 32,000 tokens) and precise analysis, Gemma 4 31B or Qwen 3.8 27B are a better fit if a computer with 32 GB is available.
How to start
ollama run gemma4:e4bWith Ollama or LM Studio on a laptop with 16 GB. The 4-bit version needs about 6 GB, leaving enough memory for other programs.
Suitable interfaces and programs from our directory: Ollama, LM Studio
Licence: Free for commercial use
The licence allows commercial use, modification and redistribution; the main obligation is to keep the licence notice.
Hugging Face ↗Similar models
Comparable models.
Google · United States
Gemma 4 E2B
The smallest Gemma 4 model, built for devices with little memory. Accepts text, images and audio.
Free for commercial use · Runs on: Laptop with 16 GB
In the overview →Mistral AI · France
Ministral 3 8B
An even smaller version of Ministral 3 with image input that gets by with 16 GB of memory.
Free for commercial use · Runs on: Laptop with 16 GB
In the overview →Alibaba (Qwen) · China
Qwen3.5 9B
Small Qwen model with image input and 262,000 tokens of context that runs on a laptop with 16 GB. Suited to summaries, drafts and simple assistant tasks.
Free for commercial use · Runs on: Laptop with 16 GB
In the overview →
FAQ
What hardware do I need for Gemma 4 E4B?
In the common 4-bit version, Gemma 4 E4B needs about 6 GB of memory including headroom for context and runtime. That fits: Laptop with 16 GB. The more accurate 8-bit version needs about 8 GB. Long documents increase the requirement.
May I use Gemma 4 E4B commercially?
The licence allows commercial use, modification and redistribution; the main obligation is to keep the licence notice.
How good is Gemma 4 E4B in German?
In EuroEval’s German comparison, Gemma 4 E4B ranks 57 of 140. There are no human ratings in German on LMArena yet; test it with your own examples.
AI consulting
Your own language model for your company?
Together we find the model that fits your tasks, your hardware and your data protection requirements, test it with real documents and set up operation.