# Fluent AI: Offline & Cloud LLM > AI chat assistant with offline models - private, customizable, multimodal Canonical page: [https://chrome-stats.com/d/com.readheights.fluentai](https://chrome-stats.com/d/com.readheights.fluentai) ## Overview - **ID:** `com.readheights.fluentai` - **Platform:** Android - **Type:** Android app - **Status:** Available - **Publisher:** ReadHeights Technologies Private Limited - **Category:** PRODUCTIVITY - **Downloads:** 21,109 - **Version:** 1.3.7 - **Last updated:** 2026-08-20 - **First published:** 2025-10-29 - **Publisher country:** IN - **Size:** 109 MB - **Data as of:** 2026-09-05 - **Store listing:** [Google Play Store](https://play.google.com/store/apps/details?id=com.readheights.fluentai) - **Website:** [https://readheights.com/](https://readheights.com/) - **Privacy policy:** [https://readheights.com/privacy](https://readheights.com/privacy) ## Description πŸ€– Fluent AI β€” Private Offline LLM + Claude, GPT-4 & Gemini

Run AI entirely on your device β€” no cloud, no account, no data sent anywhere. Then switch to Claude, GPT-4 or Gemini when you need more power. One app. Every AI. Always private.

✨ WHAT'S NEW IN v1.3

πŸ₯ MEDICAL AI (MedGemma)
β€’ Google's MedGemma 4B β€” clinical Q&A and biomedical text, 100% on-device
β€’ Requires accepting Google's Health AI Developer Foundation Terms
β€’ Not a substitute for professional medical advice

πŸ€– AGENTIC MODE
β€’ On-device AI agent with 12 built-in skills
β€’ Runs tasks autonomously: calendar events, web research, document digest, trip planning
β€’ Agent Task Inspector β€” see every reasoning step in real time
β€’ 3 free agent runs/day β€” no subscription needed to start
β€’ Scheduled tasks available with Premium

⚑ LITERT MTP β€” UP TO 2Γ— FASTER
β€’ Gemma 4n E2B/E4B with Multi-Token Prediction on Android GPU
β€’ Speculative decoding β€” more tokens per step, same quality
β€’ Tok/s display measures decode-phase speed only for accurate results

πŸ‘οΈ ON-DEVICE VISION (Android)
β€’ Attach photos using Gemma 4n β€” processed entirely on-device
β€’ No image uploaded to any server, ever

πŸ”’ PRIVACY FIRST
β€’ Conversations stay on your device
β€’ Optional local models = zero cloud data
β€’ API keys encrypted with AES β€” never stored in plain text
β€’ No mandatory account required

🧠 LOCAL AI MODELS
β€’ GGUF / llama.cpp: Gemma 3/4, Qwen 3.5, Phi-4, Llama, DeepSeek R1, Nemotron, MedGemma
β€’ LiteRT (Android GPU/NPU): Gemma 4n E2B/E4B β€” vision + MTP speculative decoding
β€’ Apple MLX: Native Metal on Apple Silicon and iOS 18+ (A17 Pro+)
β€’ Q5_K GPU acceleration on Qualcomm Adreno (alongside Q4_0)
β€’ Device-aware model recommendations based on your RAM and chipset
β€’ Browse, download, and manage models in-app β€” no sideloading needed
β€’ Import custom GGUF from HuggingFace URL or device storage

☁️ CLOUD AI
β€’ Claude (Anthropic), GPT-4 (OpenAI), Gemini (Google)
β€’ OpenRouter β€” 200+ models via a single API key
β€’ Streaming, vision, and tool calling across all providers

🌐 ONLINE SERVERS
β€’ Ollama Cloud and self-hosted Ollama
β€’ LM Studio, vLLM, LocalAI, and any OpenAI-compatible /v1 API
β€’ Multiple server profiles with per-profile encrypted auth headers

🎀 VOICE MODE
β€’ 5 conversation modes: Normal, Interview, Learning, Storytelling, Translation
β€’ Animated waveform, voice commands (speed, repeat, stop)
β€’ Quick-capture mic button directly in the chat input bar

πŸ“š KNOWLEDGE BASES (RAG)
β€’ Import PDFs, TXT, and Markdown β€” AI references your docs when answering
β€’ Semantic search for relevant context, topic and project organisation

πŸ”§ POWER FEATURES
β€’ Tool calling: Calculator, DateTime, Weather, Web Search, mem0 Memory
β€’ MCP servers: GitHub, Slack, Notion, Supabase, and 20+ presets
β€’ Code execution: Python, Bash, Node.js from code blocks (desktop + mobile JS)
β€’ Model benchmarking: tok/s, TTFT, MMLU-50 quality score, shareable PNG cards
β€’ Slash commands: /agent, /clear, /export, /voice, /template and more
β€’ Per-chat thinking toggle for Qwen3, DeepSeek R1, Nemotron reasoning models
β€’ URL context injection β€” paste a link, AI reads the page for context
β€’ Polish Before Send β€” AI rewrites your draft before you hit send
β€’ Continue button β€” resumes responses cut off at the token limit

πŸ“ CHAT ORGANISATION
β€’ Folders, tags, and cross-chat full-text search across every message
β€’ HuggingFace model browser with bookmarks and memory fitness badges
β€’ Conversation branching and message reactions

🌟 PREMIUM (OPTIONAL)
β€’ Ad-free experience
β€’ Scheduled agent tasks (recurring or one-time)
β€’ Priority feature access and advanced analytics

πŸ“± PERFECT FOR
βœ“ Privacy-focused users β€” local models, zero cloud data
βœ“ Android power users β€” LiteRT GPU/NPU with MTP acceleration
βœ“ Developers β€” benchmark GGUF, LiteRT, and MLX side-by-side
βœ“ Healthcare researchers β€” MedGemma on-device, no upload needed
βœ“ Students β€” knowledge bases for study documents and materials
βœ“ Professionals β€” agentic tasks, document Q&A, and tool calling ## Rankings - #423,911 β€” Overall - #30 β€” возмоТности ai ## Permissions and access ### Permissions - `android.permission.ACCESS_ADSERVICES_AD_ID` - `android.permission.ACCESS_ADSERVICES_ATTRIBUTION` - `android.permission.ACCESS_ADSERVICES_TOPICS` - `android.permission.ACCESS_NETWORK_STATE` - `android.permission.ACCESS_NOTIFICATION_POLICY` - `android.permission.BLUETOOTH` - `android.permission.BLUETOOTH_ADMIN` - `android.permission.BLUETOOTH_CONNECT` - `android.permission.FOREGROUND_SERVICE` - `android.permission.FOREGROUND_SERVICE_DATA_SYNC` - `android.permission.INTERNET` - `android.permission.POST_NOTIFICATIONS` - `android.permission.RECEIVE_BOOT_COMPLETED` - `android.permission.RECORD_AUDIO` - `android.permission.REQUEST_IGNORE_BATTERY_OPTIMIZATIONS` - `android.permission.VIBRATE` - `android.permission.WAKE_LOCK` - `com.android.vending.BILLING` - `com.android.vending.CHECK_LICENSE` - `com.google.android.c2dm.permission.RECEIVE` - `com.google.android.finsky.permission.BIND_GET_INSTALL_REFERRER_SERVICE` - `com.google.android.gms.permission.AD_ID` - `com.google.android.providers.gsf.permission.READ_GSERVICES` - `com.readheights.fluentai.DYNAMIC_RECEIVER_NOT_EXPORTED_PERMISSION` ## Similar extensions and apps - [Private AI](https://chrome-stats.com/d/us.valkon.privateai) β€” 90,771 users, 3.90 / 5 - [Offline AI Chat - Local LLM](https://chrome-stats.com/d/com.freeai.chat) β€” 14,672 users, 2.40 / 5 - [MultiAI | Online & Offline AI](https://chrome-stats.com/d/com.matrix.multigpt) β€” 1,752 users - [Google AI Edge Gallery](https://chrome-stats.com/d/com.google.ai.edge.gallery) β€” 2,094,607 users - [PROLINE AI](https://chrome-stats.com/d/com.prolineweb.prolineai) β€” 174 users - [Gemma 4- Offline AI](https://chrome-stats.com/d/com.offlinesecureai.android) β€” 830 users - [Gemmi Chat: Offline AI chatbot](https://chrome-stats.com/d/com.sensornotes.gemmi.ai) β€” 7,024 users - [Nexus AI: Private Offline LLM](https://chrome-stats.com/d/nexus.ai.offline) β€” 309 users - [InferrLM: On-device AI](https://chrome-stats.com/d/com.gorai.ragionare) β€” 932 users, 3.20 / 5 - [Local LLM](https://chrome-stats.com/d/com.AlexPechkin.LocalLLM) β€” 1,348 users - [LocalGPT: Offline AI Assistant](https://chrome-stats.com/d/com.rnllamaexample) β€” 6,908 users - [Pocket AI Offline Assistant](https://chrome-stats.com/d/com.hectasquare.pocketAI) β€” 1,608 users --- Source: [Chrome-Stats](https://chrome-stats.com/d/com.readheights.fluentai)