Multiple model sizes
Choose lightweight models for older phones or stronger models for capable devices.
Q17 Private AI combines local models, voice, vision, image generation, projects, agents, runtime selection, and optional web routing inside one Android-first AI workspace.
The app acts like a private AI workspace where each task can move through the right local or optional online path.
Text, voice, image, camera, PDF, XML, or other supported attachments.
Local chat model, vision model, image engine, voice tool, or optional web support.
Chats, projects, agents, histories, model manager, and local app storage.
Answer, voice reply, image result, workflow result, or saved conversation.
Use supported on-device models for writing, summaries, reasoning, planning, coding-style help, and daily assistant tasks.
Choose lightweight models for older phones or stronger models for capable devices.
Start with free models and unlock premium local models when needed.
Chats are designed around local-first usage instead of cloud-first dependency.
Download, install, activate, delete, and manage models inside the app.
Q17 Private AI is built around a practical mobile voice workflow with speech input, offline voice output options, and downloadable voice packs.
Ask with your voice and hear spoken responses.
Use offline voice packs instead of forcing one fixed cloud voice.
Interrupt or continue conversations depending on your voice mode settings.
Pair multilingual models with multilingual voice tools where supported.
Q17 Private AI supports image and document-style workflows for users who want more than plain text chat.
Ask questions about images and screenshots using supported vision models.
Capture images from your phone and send them into Q17 Private AI.
Use text, PDF, XML, and document attachments in supported flows.
Q17 Private AI routes text, image, and attachment tasks toward the right model path.
Q17 Image Studio is the on-device image-generation workspace inside Q17 Private AI. Install a supported local pack, then generate directly on your Android device.
Create images from prompts directly inside Q17 Image Studio after installing a supported pack.
Current Image Studio packs are Realistic Vision, DreamShaper 8, and Counterfeit V2.5.
Fast and Balanced use LCM paths for quicker on-device generation.
Quality and Detailed use DPM paths and are currently experimental.
Q17 Private AI gives users organized spaces for serious work, repeated tasks, and task-focused assistant workflows.
Keep related chats, files, and work grouped together.
Use task-focused assistants for specific workflows.
Keep normal chat, project chat, and agent work cleaner.
Organize serious work without losing context inside one long chat.
Q17 Private AI is designed around practical mobile AI performance across different device classes and backend capabilities.
Backend-aware runtime with CPU fallback, guarded OpenCL/Vulkan acceleration, and planned Hexagon/NPU support.
Q17 Private AI can fall back when a backend is unavailable or unsafe for a model.
Lightweight models help older and budget phones still run useful AI.
Model and voice downloads are handled with install state, progress, retry, and verification logic.
The app keeps web support as a deliberate mode instead of making every answer cloud-dependent.
Web support can be enabled or disabled from the chat workflow.
Use local models as the workspace while optional online context can support current questions.
The UI is designed to show when web support is being used.
Offline and local-first usage remains the core product direction.
Q17 Private AI is built around local storage, model files, downloadable voices, image packs, and install-state checks.
Downloaded models and resources are managed on the phone.
The app tracks download, install, retry, and activation states.
Users can choose text, vision, voice, and Q17 Image Studio packs where supported.
Smaller or stronger packs can fit different Android device classes.
Q17 Private AI can expose stronger models and advanced local packs while keeping the same private Android workspace.
Premium access can unlock selected model packs or advanced capabilities.
Premium purchase flows are expected to use Google Play Billing.
Free and premium models live inside the same chat, voice, vision, and project system.
Model availability and runtime support can depend on device storage, RAM, and backend support.
Q17 Image Studio currently supports Realistic Vision, DreamShaper 8, and Counterfeit V2.5. Fast and Balanced use LCM paths. Quality and Detailed use DPM paths and are currently experimental.
Built for more realistic and photoreal-style image generation on supported Android devices.
Designed for stylized, concept-art, fantasy, and more creative visual outputs.
Suited for anime-inspired and illustrated image generation styles.
Open the official Q17 Private AI Play Store listing to install the Android app.