Released April 2, 2026 — Apache 2.0

Your Community Deserves
Its Own AI

Gemma 4 by Google DeepMind brings frontier-level intelligence to local hardware. No subscriptions. No data leaving your network. No limits. Build a shared AI workbench that serves everyone — privately, freely, forever.

4Model Sizes
140+Languages
256KContext Window
€0License Cost

Four Models, Every Scale

From a phone in your pocket to a workstation under your desk — one family covers it all.

📱

E2B — Ultra Edge

2.3B effective params. Runs on phones, Raspberry Pi, any 2GB+ device. Text + Image + Audio. 128K context. The smallest model with native speech recognition.

💻

E4B — Edge Pro

4.5B effective params. Laptops and tablets with 8GB RAM. Text + Image + Audio. 128K context. Complex tasks without internet.

⚡

26B A4B — The MoE Star

26B total, only 3.8B active per token. Runs at 4B speed, thinks at 30B quality. 256K context. The efficiency breakthrough — best quality-per-watt in open AI.

🖥️

31B Dense — Maximum

31B params, full dense architecture. Top quality for demanding workloads. 256K context. #3 open model globally on LMArena.

What Does It Actually Cost to Build?

Three practical build tiers — from a repurposed old PC to a dedicated AI workstation.

🟢
BASE — Entry Server
€200 – €500
GPUGTX 1080 8GB (used)
RAM16–32 GB
ModelE4B / 26B (Q2)
Speed~10–20 tok/s

Runs E4B at full speed or 26B MoE at Q2 quantization. Good enough for a small team doing document analysis, translation, and basic coding help. Could be a repurposed gaming PC or office desktop with a used GPU added.

🔴
TOP — AI Workstation
€2,500 – €5,000
GPURTX 5090 32GB / 2×3090
RAM64–128 GB
Model31B Dense (Q6–Q8)
Speed~50–85 tok/s

Maximum quality. 31B Dense at high quantization with full 256K context. Serves 5–10 concurrent users. Professional-grade coding, research, document intelligence. Can fine-tune models on your own data. Multi-GPU for heavy workloads.

Not Just Open — Best-in-Class

Gemma 4 doesn't just compete. It leads reasoning, math, and coding among all open models.

Gemma 4 31B vs Gemma 3 27B — Generational Leap

MMLU Pro (knowledge)85.2% vs 67.6%
AIME 2026 (math reasoning)89.2% vs 20.8%
LiveCodeBench v6 (coding)80% vs 29.1%
Codeforces ELO2150 vs 110

vs Competition (April 2026)

FeatureGemma 4 31BGemma 4 26B MoEQwen 3.5 27BLlama 4 Scout
LMArena Score1452 (#3)1441 (#6)~1420~1400
Active Params31B3.8B27B17B
Context256K256K256K10M
Languages140+140+201~100
Edge Models✓ E2B/E4B with audio✓ small variants✗ min 109B
LicenseApache 2.0Apache 2.0Apache 2.0Restricted
GPU Needed (Q4)~20 GB~16 GB~18 GB~65 GB

What Can a Shared AI Server Do?

A local Gemma 4 workbench becomes infrastructure — like shared internet or a community library.

01

Multilingual Document Hub

Translate contracts, forms, and official documents across 140+ languages instantly. No uploading sensitive personal data to cloud services. Ideal for immigrant support, rural offices, NGOs.

E4B edge26B server
02

Local Business Assistant

Pricing research, inventory management, customer email drafting, invoice processing, social media content — in your language, with your context. Small shops and cooperatives get AI tools without monthly SaaS fees.

26B server
03

Education & Tutoring

Personalized tutoring for students of all ages. Math reasoning (89% AIME), science explanations, language learning, homework help. Works offline — perfect for areas with limited internet.

E4B tablet26B server
04

Code Workshop & Maker Lab

Codeforces ELO 2150 — competitive programmer level. Debug code, build apps, learn programming, automate workflows. Native tool calling enables agentic coding: the AI writes, tests, and fixes code autonomously.

26B server31B workstation
05

Agricultural & Land Management

Analyze soil reports, identify plants from photos, plan crop rotations, process drone imagery, interpret weather data. Vision capabilities handle charts, maps, and field photos natively.

E4B phone26B server
06

Health Information Access

Understand medical reports, translate prescriptions, access health knowledge in native languages. Not a replacement for doctors — but a bridge for underserved communities to understand their own medical documents.

26B server
07

Creative & Cultural Projects

Write stories, scripts, poetry in any language. Analyze historical documents. Transcribe oral histories via audio input. Generate educational materials for cultural preservation.

E2B audio26B server
08

Privacy-First Legal Aid

Summarize legal documents, explain regulations, draft simple contracts. All data stays local — critical for sensitive legal matters. Supports 140+ languages for diverse communities.

26B server31B workstation

Cloud vs. Local — The Math

A one-time hardware investment replaces perpetual subscription costs. For a team of 10 people.

☁️ Cloud APIs

€200+/mo

ChatGPT Team: €25/user × 10 = €250/mo
Claude Pro: €20/user × 10 = €200/mo
API usage: variable, can spike
= €2,400–3,000+/year
+ your data leaves your network

🏠 Local Gemma 4

€0/mo

One-time hardware: €800–1,500
Electricity: ~€10–20/mo
Software: free (Ollama + open tools)
= €0 after month 5–7
+ total data privacy & sovereignty

Break-even in ~6 months. Free forever after.

Every month after break-even, your community saves €200–300 that stays in the local economy.

One Command. Infinite Potential.

Install Ollama on any Linux, Mac, or Windows machine. Download Gemma 4. Start serving your community today.

ollama run gemma4

Apache 2.0 licensed — no restrictions, no user limits, no usage policies.
Supported from day one by: Ollama, llama.cpp, LM Studio, vLLM, Hugging Face, MLX, NVIDIA NIM.