Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking
by DavidAU · built on gemma 3 4b heretic uncensored abliterated Extreme · rank #100 of 135
Benchmark scores
This model is not ranked on the UGI Leaderboard yet. Estimated speed: ~12–22 tok/s (estimate), a rough range for a Q4_K_M file of this size on a recent flagship chip (Snapdragon 8 Gen 3 or A17 Pro class). It is not a measurement from this model.
About this model
Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking is a 4B-parameter language model by DavidAU, derived from DavidAU/gemma-3-4b-it-heretic-uncensored-abliterated-Extreme using Heretic, an automated abliteration tool. You can download it and run it on your phone for free.
Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking is a 4B model by DavidAU of gemma 3 4b heretic uncensored abliterated Extreme, using heretic, an automated abliteration tool. It can read images and writes out its reasoning before it answers. The Q4_K_M file is 2.49 GB. It writes out its reasoning before answering, which is slower but more careful, and accepts images, so it can read a screenshot of a chart or document.
Technically, it uses the gemma3 architecture, supports a context window of 131,072 tokens (a phone will use far less).
The device-ready GGUF files were produced by Andycurrent.
Which RAM sizes can run it?
| Device RAM | iOS-safe | Android-safe |
|---|---|---|
| 4 GB RAM | Fits with Q3_K_M (~2.1 GB) | Fits with Q4_K_M (~2.49 GB) |
| 6 GB RAM | Fits with Q4_K_M (~2.49 GB) | Fits with Q4_K_M (~2.49 GB) |
| 8 GB RAM | Fits with Q4_K_M (~2.49 GB) | Fits with Q4_K_M (~2.49 GB) |
| 12 GB RAM | Fits with Q4_K_M (~2.49 GB) | Fits with Q4_K_M (~2.49 GB) |
| 16 GB+ RAM | Fits with Q4_K_M (~2.49 GB) | Fits with Q4_K_M (~2.49 GB) |
iOS-safe minimum: 4 GB (using Q3_K_M). Android-safe minimum: 4 GB. On iOS we treat about 57% of device RAM as usable (in the 55–60% range), because the system keeps a large share and may evict an app that goes higher. On Android we treat about 67% of device RAM as usable (in the 65–70% range), leaving room for the system and the app. See the guide.
Run it on your phone
Compatible apps:
iOS
- PocketPal AI — Free on the App Store. Can search Hugging Face and load a GGUF file on the device.
- Layla — On the App Store as a paid app. Load a custom GGUF from files on the device.
Android
- PocketPal AI — Free on Google Play. Can search Hugging Face and load a GGUF file on the device.
- Layla — On Google Play. Load a custom GGUF from storage. A limited free APK is also offered on the developer site.
- ChatterUI — Free and open source. Install the APK from the GitHub releases page. iOS is not available.
Hugging Face repository ID:
Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUFRecommended file: Q4_K_M — Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q4_k_m.gguf (about 2.49 GB)
- Install one of the apps above for your operating system. PocketPal AI is free on both iOS and Android and can search Hugging Face directly.
- In the app, add a model from Hugging Face and paste this repository ID:
Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF - Download the Q4_K_M file (
Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q4_k_m.gguf) (about 2.49 GB), or Q3_K_M (about 2.1 GB) if memory is tight. Use wifi. This is the only step that needs internet. - Load the file and start chatting. Switch on airplane mode if you like: it keeps working. A first prompt to try:
I'm considering a decision this week. Give me the strongest case for it, the strongest case against it, and the one fact that would change my mind. Keep it short.
If the app is killed or feels slow, close other apps, try the smaller file, or pick a model with a lower iOS-safe or Android-safe minimum. On a computer, the same files work in LM Studio, Ollama and llama.cpp.
Available GGUF files (7)
| Quantization | Size | Download |
|---|---|---|
| Q2_K Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q2_k.gguf | 1.73 GB | Download |
| Q3_K_MSMALLER Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q3_k_m.gguf | 2.10 GB | Download |
| Q4_K_MRECOMMENDED Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q4_k_m.gguf | 2.49 GB | Download |
| Q5_K_M Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q5_k_m.gguf | 2.83 GB | Download |
| Q6_K Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q6_k.gguf | 3.19 GB | Download |
| Q8_0 Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_Q8_0.gguf | 4.13 GB | Download |
| F16 Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_F16.gguf | 7.77 GB | Download |
| Vision projector Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_mmproj_f16.gguf | 851 MB | Download |
Direct downloads from Andycurrent/Gemma-3-4B-VL-it-Gemini-Pro-Heretic-Uncensored-Thinking_GGUF on Hugging Face. Lower Q numbers are smaller and slightly less accurate. Vision needs the projector (mmproj) file alongside the model.