Frequently Asked Questions

Everything you need to know about AI Troll: privacy, model support, memory guidelines, and local performance.

General Questions

What is AI Troll?

AI Troll is a mobile app that lets you run AI language models directly on your phone or tablet. All processing happens locally on your device—no internet connection required after downloading a model.

Is AI Troll free?

Yes, AI Troll is free to download and use. Licenses vary by model; you must review and comply with each model's specific license on HuggingFace before downloading.

Does AI Troll require an internet connection?

Only for downloading AI models. Once a model is downloaded, you can use AI Troll completely offline.

Does AI Troll support multiple languages?

Yes! Since version 2.0, AI Troll has been fully localized into 12 major languages, including English, Arabic, German, Spanish, French, Italian, Japanese, Korean, Portuguese, Russian, Turkish, and Simplified Chinese. You can switch the app language anytime in the Settings menu.

Can AI Troll translate text?

Yes — if your active model is multilingual. When AI Troll detects a translation-capable model (roughly 1B parameters or larger, e.g. Qwen or Gemma instruct models), a translator button appears in the chat toolbar. It opens a dedicated page where you pick a language pair (or use auto-detect) and translate entirely offline — nothing is sent anywhere and translations are not saved. Quality depends on the model and the language pair; do not rely on it for legal, medical, or safety-critical content. The feature can be turned off under Settings → Features.

Privacy & Security

Is my data private?

Yes! AI Troll is designed with privacy as the top priority. All conversations happen locally on your device. No data is collected, no analytics, no tracking. Your conversations never leave your phone.

Does AI Troll send my conversations to the cloud?

No. All AI processing happens 100% on your device. Your conversations are stored only in local app storage and are never transmitted to any server.

Do I need to create an account?

No. AI Troll does not require any account, registration, or sign-in.

What data does AI Troll collect?

None. AI Troll does not collect any personal data, usage analytics, or any other information.

Is there an age requirement?

Yes. AI Troll is for adults only (18+). It can run uncensored third-party models that may produce mature content. In regions where Google Play provides age signals (as required by law), AI Troll uses the Google Play Age Signals API to block users identified as being under 18; elsewhere we rely on your confirmation that you are 18 or older. The check happens through Google Play on your device and never reveals your date of birth.

Models & Downloads

Where do the AI models come from?

AI Troll downloads models from HuggingFace, a popular repository of open-source AI models. You can also import your own GGUF models from your device storage.

What model format does AI Troll support?

AI Troll supports GGUF (GPT-Generated Unified Format) models. This format is optimized for efficient inference and includes everything needed to run the model.

How big are the AI models?

Model sizes vary from several hundreds of MB to 10GB or more, depending on the model architecture and quantization level. Smaller quantized versions (Q4, Q5) are recommended for mobile devices.

Which model should I download?

For most phones, we recommend starting with smaller quantized models (Q4_K_M or Q5_K_M) in the 1-3GB range. These offer a good balance between quality and performance.

Can I delete downloaded models?

Yes. Go to Model Management and tap the delete option for any model you want to remove. This will free up storage space on your device.

Can I use my own models?

Yes! You can import any GGUF model file from your device storage using the "Import Local Model" feature.

Does AI Troll support Google's Gemma or Meta's LLaMA models?

Yes! AI Troll natively supports popular open-source model families like Google's Gemma, Meta's LLaMA, Microsoft's Phi, and Alibaba's Qwen. The application automatically detects the correct chat template and configures appropriate context length limits to guarantee smooth inference on your hardware.

Why is the HuggingFace model list paginated?

**In versions prior to 2.0,** loading the entire compatible model list at once could cause search lag or UI freezing on low-end devices. **The pagination mechanism introduced in version 2.0** loads models in clean, fast batches (pages), enabling you to search and filter through hundreds of GGUF models smoothly and efficiently.

Performance & RAM Requirements

Why is the AI responding slowly?

AI processing is computationally intensive. Larger models require more processing time. Try using a smaller/more quantized model, closing background apps to free up RAM, and ensuring your device isn't overheating.

How much storage do I need?

This depends on the models you download. A typical setup with 2-3 models might require 3-10GB of storage.

Does AI Troll drain battery?

AI processing is CPU-intensive, which does use more battery than typical apps. For extended sessions, we recommend keeping your device plugged in.

How much RAM does AI Troll need?

RAM requirements depend directly on the model file size and quantization compression type:

By Model File Size

1-2GB models: 4GB RAM minimum, 6GB+ recommended

2-4GB models: 6GB RAM minimum, 8GB+ recommended

4-6GB models: 8GB RAM minimum, 10GB+ recommended

6GB+ models: 12GB+ RAM (high-end only)

By Quantization Type

Q4_K_M (high compress): ~1-3GB files → 4-6GB RAM

Q5_K_M (balanced): ~2-4GB files → 6-8GB RAM

Q6_K (light compress): ~3-5GB files → 8-10GB RAM

Q8_0 (no compress): ~4-8GB files → 8GB+ RAM

Device Quick Guide

4-6GB RAM Phone: GGUF models under 2GB (Q4_K_M)

6-8GB RAM Phone: 2-4GB models (Q4_K_M or Q5_K_M)

8GB RAM Phone: Models under 4GB (Max 50% RAM logic)

12GB+ RAM Phone: Most mobile GGUFs (up to Q8_0)

Tip: The app automatically checks system memory and warns you before loading a model that's too large for your device. Close other running apps to free up RAM before loading large models.

Example: A standard "Llama-3.2-3B-Q4_K_M.gguf" file (2.1GB) needs approximately 4-6GB RAM to run smoothly.

AI Responses & Settings

Can I trust the AI's responses?

AI responses should be treated as suggestions, not facts. AI models can generate inaccurate, incomplete, or misleading information. Always verify important information from reliable sources.

Is the AI suitable for professional advice?

No. You should NOT rely on AI responses for medical, legal, financial, or other professional advice. Always consult qualified professionals for such matters.

Why does the AI sometimes give wrong answers?

AI language models work by predicting the most likely next words based on their training. They don't actually "know" facts—they generate plausible-sounding text. This means they can confidently state incorrect information.

Can I customize how the AI responds?

Yes! In Settings, you can fully adjust:

  • Context Limit: How much conversation history the AI can remember (based on your device's RAM).
  • Response Length: Maximum tokens per AI response.
  • Temperature: How creative/random vs. focused the responses are.
  • Top-K/Top-P: How many token choices the AI considers.
  • Repetition Penalty: Reduces repetitive sentences in responses.
What if the AI generates inappropriate content?

If the AI produces content you find inappropriate or offensive, you can flag and hide it directly in the chat. Long-press on any AI message to access the flag option. Flagged content will be hidden from view.

Troubleshooting

The app crashes when loading a model. What should I do?

The model may be too large for your device's available memory. Try:

  • Close other running apps to free up system RAM
  • Download a smaller or more compressed (quantized) version of the model
  • Restart your device and try again
My download keeps failing. How can I fix it?

Try these steps:

  • Check your internet connection
  • Ensure you have enough free storage space on your device. WARNING: Manually verify you have enough free storage space *before* initiating downloads.
  • Check if "WiFi Only Downloads" is enabled in Settings (downloads won't start on mobile data)
  • Try a smaller model first to verify connection and storage permissions
  • Note: Downloads continue in the background even if you close the app—you'll get a notification when complete.
How do I clear my chat history?

You can always delete chats in the list of chats inside the app, or clear the app's data completely from Android Settings.