Google
Gemma 4 QAT E2B
Google’s compact assistant for writing and everyday questions. E means effective parameters; the full download is larger. Text chat only.
About Gemma 4 QAT E2B
Gemma 4 QAT E2B is Google's quantization-aware-trained checkpoint of Gemma 4 E2B, its on-device-focused model. Because quantization is simulated during training instead of applied afterwards, the Q4_0 file keeps quality closer to full precision. E2B counts effective parameters, so the full download is larger: about 3.4 GB.
Every model in this catalog runs locally in the Offline AI Chat app. After the download finishes, Gemma 4 QAT answers on your phone with no connection, and your prompts and replies are never uploaded.
- Download size
- 3.3 GB (Q4_0 GGUF)
- Memory needed
- 6 GB minimum · 8 GB recommended
- Context in app
- 2,048 tokens
- Runs on
- Android 8+, 64-bit · iOS in private testing
- Access
- Offline Plus subscription
- License
- Apache-2.0
- Model source
- Google on Hugging Face