diff --git a/yuy-chat-complete/.gitignore b/yuy-chat/.gitignore
similarity index 100%
rename from yuy-chat-complete/.gitignore
rename to yuy-chat/.gitignore
diff --git a/yuy-chat-complete/Cargo.toml b/yuy-chat/Cargo.toml
similarity index 100%
rename from yuy-chat-complete/Cargo.toml
rename to yuy-chat/Cargo.toml
diff --git a/yuy-chat/README.md b/yuy-chat/README.md
new file mode 100644
index 0000000..696ecd9
--- /dev/null
+++ b/yuy-chat/README.md
@@ -0,0 +1,180 @@
+# yuy-chat
+
+<div align="center">
+
+```
+$$\     $$\                    
+\$$\   $$  |                   
+ \$$\ $$  /$$\   $$\ $$\   $$\ 
+  \$$$$  / $$ |  $$ |$$ |  $$ |
+   \$$  /  $$ |  $$ |$$ |  $$ |
+    $$ |   $$ |  $$ |$$ |  $$ |
+    $$ |   \$$$$$$  |\$$$$$$$ |
+    \__|    \______/  \____$$ |
+                     $$\   $$ |
+                     \$$$$$$  |
+                      \______/ 
+```
+
+**Beautiful TUI chat interface for local AI models**
+
+[![Rust](https://img.shields.io/badge/rust-1.70%2B-orange.svg)](https://www.rust-lang.org)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+
+</div>
+
+---
+
+## 🌟 Features
+
+- ✨ **Beautiful TUI** - Gorgeous terminal interface powered by ratatui
+- 🔍 **Auto-discovery** - Automatically finds `.gguf` and `.llamafile` models
+- 🎨 **Presets** - Creative, Balanced, and Precise modes
+- 💾 **Save conversations** - Keep your chat history
+- 🌐 **HuggingFace API** - Use models from HuggingFace (optional)
+- ⚡ **Fast & Lightweight** - ~5MB binary, minimal dependencies
+- 🚀 **Streaming responses** - See words appear as they're generated
+- 🎯 **Zero configuration** - Just run and chat
+
+## 📦 Installation
+
+### From source:
+
+```bash
+git clone https://github.com/YuuKi-OS/yuy-chat
+cd yuy-chat
+cargo build --release
+```
+
+### Install globally:
+
+```bash
+cargo install --path .
+```
+
+## 🚀 Quick Start
+
+```bash
+# Run yuy-chat
+yuy-chat
+
+# It will auto-scan ~/.yuuki/models/ for .gguf and .llamafile files
+# Select a model and start chatting!
+```
+
+## 📁 Supported Model Formats
+
+- ✅ **GGUF** (`.gguf`) - Runs with llama.cpp
+- ✅ **Llamafile** (`.llamafile`) - Self-contained executables
+
+## 🎮 Controls
+
+### Model Selector
+- `↑/↓` or `j/k` - Navigate models
+- `Enter` - Select model
+- `R` - Refresh model list
+- `Q` - Quit
+
+### Chat
+- `Type` - Write your message
+- `Enter` - Send message
+- `Shift+Enter` - New line
+- `Ctrl+Enter` - Send (always)
+- `Ctrl+C` - Open menu
+- `Ctrl+L` - Clear chat
+- `Ctrl+S` - Save conversation
+- `↑/↓` - Scroll chat (when input is empty)
+
+### Menu
+- `1` - Change model
+- `2` - Change preset
+- `3` - Save conversation
+- `4` - Load conversation
+- `5` - Clear chat
+- `6` - Settings
+- `Q` - Back to chat
+
+## ⚙️ Configuration
+
+Config file location: `~/.config/yuy-chat/config.toml`
+
+```toml
+models_dir = "/home/user/.yuuki/models"
+hf_token = "hf_xxxxxxxxxxxxx"  # Optional
+default_preset = "Balanced"
+save_history = true
+theme = "Dark"
+```
+
+## 🎯 Presets
+
+- **Creative** (temp: 0.8, top_p: 0.9) - More random and creative
+- **Balanced** (temp: 0.6, top_p: 0.7) - Good middle ground
+- **Precise** (temp: 0.3, top_p: 0.5) - More focused and deterministic
+
+## 🌐 HuggingFace Integration
+
+Add your HuggingFace token in settings to use models via API:
+
+1. Press `Ctrl+C` → `6` (Settings)
+2. Edit `HuggingFace Token`
+3. Paste your token from https://huggingface.co/settings/tokens
+4. Save and refresh models
+
+## 📚 Directory Structure
+
+```
+~/.config/yuy-chat/
+├── config.toml              # Configuration
+└── conversations/           # Saved chats
+    ├── conversation-20240206-143022.json
+    └── conversation-20240206-150133.json
+```
+
+## 🔧 Requirements
+
+- **Rust 1.70+** (for building)
+- **llama.cpp** (for .gguf models) - Install with: `yuy runtime install llama-cpp`
+- **chmod +x** (for .llamafile models)
+
+## 🤝 Integration with yuy
+
+yuy-chat is designed to work alongside [yuy](https://github.com/YuuKi-OS/yuy):
+
+```bash
+# Download models with yuy
+yuy download Yuuki-best
+
+# Chat with yuy-chat
+yuy-chat
+```
+
+## 🐛 Troubleshooting
+
+**No models found?**
+- Make sure you have models in `~/.yuuki/models/`
+- Or specify custom directory: `yuy-chat --models-dir /path/to/models`
+
+**llama.cpp not found?**
+- Install with: `yuy runtime install llama-cpp`
+- Or: `brew install llama.cpp` (macOS)
+- Or: `pkg install llama-cpp` (Termux)
+
+**Streaming not working?**
+- Ensure llama.cpp is installed and in PATH
+- Check model file permissions
+
+## 📝 License
+
+MIT License - see [LICENSE](LICENSE) file
+
+## 🌸 Credits
+
+Made with love by the Yuuki team
+
+- TUI Framework: [ratatui](https://github.com/ratatui-org/ratatui)
+- Inference: [llama.cpp](https://github.com/ggerganov/llama.cpp)
+
+---
+
+**For model management, see [yuy](https://github.com/YuuKi-OS/yuy)**
diff --git a/yuy-chat/USAGE.md b/yuy-chat/USAGE.md
new file mode 100644
index 0000000..1f728e2
--- /dev/null
+++ b/yuy-chat/USAGE.md
@@ -0,0 +1,495 @@
+# yuy-chat - Guía de Uso Completa
+
+## 📖 Contenido
+
+1. [Instalación](#instalación)
+2. [Primera Vez](#primera-vez)
+3. [Uso Diario](#uso-diario)
+4. [Configuración Avanzada](#configuración-avanzada)
+5. [Integración con HuggingFace](#integración-con-huggingface)
+6. [Tips y Trucos](#tips-y-trucos)
+7. [Troubleshooting](#troubleshooting)
+
+---
+
+## 🔧 Instalación
+
+### Termux (Android)
+
+```bash
+# Instalar Rust
+pkg install rust
+
+# Clonar y compilar
+git clone https://github.com/YuuKi-OS/yuy-chat
+cd yuy-chat
+cargo build --release -j 1  # Usar 1 thread en Termux
+
+# Instalar globalmente
+cargo install --path .
+```
+
+### Linux/macOS
+
+```bash
+# Clonar y compilar
+git clone https://github.com/YuuKi-OS/yuy-chat
+cd yuy-chat
+cargo build --release
+
+# Instalar
+cargo install --path .
+```
+
+### Windows
+
+```bash
+# Mismo proceso que Linux/macOS
+git clone https://github.com/YuuKi-OS/yuy-chat
+cd yuy-chat
+cargo build --release
+cargo install --path .
+```
+
+---
+
+## 🎬 Primera Vez
+
+### 1. Asegúrate de tener modelos
+
+yuy-chat busca modelos en `~/.yuuki/models/` por defecto.
+
+**Opción A: Usar yuy**
+```bash
+yuy download Yuuki-best
+```
+
+**Opción B: Copiar manualmente**
+```bash
+mkdir -p ~/.yuuki/models/
+cp /path/to/your/model.gguf ~/.yuuki/models/
+```
+
+### 2. Instalar llama.cpp
+
+**Termux:**
+```bash
+pkg install llama-cpp
+```
+
+**macOS:**
+```bash
+brew install llama.cpp
+```
+
+**Linux:**
+```bash
+# Descargar desde releases
+wget https://github.com/ggerganov/llama.cpp/releases/...
+chmod +x llama-cli
+sudo mv llama-cli /usr/local/bin/
+```
+
+### 3. Ejecutar yuy-chat
+
+```bash
+yuy-chat
+```
+
+Verás el selector de modelos. Usa `↑/↓` para navegar y `Enter` para seleccionar.
+
+---
+
+## 💬 Uso Diario
+
+### Flujo Básico
+
+```
+1. Ejecuta: yuy-chat
+   ↓
+2. Selecciona modelo con ↑/↓ y Enter
+   ↓
+3. Escribe tu mensaje
+   ↓
+4. Presiona Enter para enviar
+   ↓
+5. Yuuki responde (streaming)
+   ↓
+6. Continúa la conversación
+```
+
+### Atajos de Teclado Útiles
+
+**En chat:**
+- `Enter` - Enviar mensaje
+- `Shift+Enter` - Nueva línea (para mensajes multi-línea)
+- `Ctrl+L` - Limpiar chat
+- `Ctrl+S` - Guardar conversación
+- `Ctrl+C` - Abrir menú
+
+**Escribir código:**
+```
+You: Dame un ejemplo de código Python
+
+[Shift+Enter para nueva línea]
+def hello():
+    print("Hola")
+[Shift+Enter]
+
+hello()
+
+[Ctrl+Enter para enviar]
+```
+
+### Cambiar Preset
+
+```
+1. Ctrl+C (abrir menú)
+   ↓
+2. Presiona 2 (Change Preset)
+   ↓
+   Cicla entre: Creative → Balanced → Precise
+```
+
+**Cuándo usar cada preset:**
+- **Creative**: Escribir historias, brainstorming, ideas
+- **Balanced**: Uso general, conversación
+- **Precise**: Código, matemáticas, datos exactos
+
+---
+
+## ⚙️ Configuración Avanzada
+
+### Cambiar Directorio de Modelos
+
+**Método 1: Configuración**
+```bash
+yuy-chat
+Ctrl+C → 6 (Settings)
+Editar "Models Directory"
+```
+
+**Método 2: Archivo config**
+```bash
+nano ~/.config/yuy-chat/config.toml
+```
+
+```toml
+models_dir = "/custom/path/to/models"
+```
+
+### Personalizar Presets
+
+Edita el código o usa parámetros de llama.cpp directamente:
+
+```bash
+# En models/runtime.rs, modifica:
+pub fn temperature(&self) -> f32 {
+    match self {
+        Preset::Creative => 0.9,  // Más aleatorio
+        // ...
+    }
+}
+```
+
+### Tema Claro
+
+```toml
+theme = "Light"
+```
+
+---
+
+## 🌐 Integración con HuggingFace
+
+### 1. Obtener Token
+
+1. Ve a https://huggingface.co/settings/tokens
+2. Click "Create new token"
+3. Tipo: "Read"
+4. Copia el token
+
+### 2. Configurar en yuy-chat
+
+**Método A: UI**
+```
+Ctrl+C → 6 (Settings)
+Navigate to "HuggingFace Token"
+Enter → Pega tu token
+```
+
+**Método B: Config file**
+```toml
+hf_token = "hf_abcdefghijklmnopqrstuvwxyz1234567890"
+```
+
+### 3. Usar Modelos de HF
+
+Después de configurar el token:
+```
+yuy-chat
+[Verás modelos locales + modelos HF API]
+
+> Yuuki-best.gguf (Local)
+  Yuuki-3.7.gguf (Local)  
+  Yuuki-best (HF API) <-- Usa la API
+```
+
+**Ventajas:**
+- No ocupa espacio local
+- Siempre actualizado
+- Acceso a modelos privados
+
+**Desventajas:**
+- Requiere internet
+- Más lento que local
+- Rate limits en plan gratis
+
+---
+
+## 💡 Tips y Trucos
+
+### Guardar Conversaciones Importantes
+
+```
+Ctrl+S mientras chateas
+→ Se guarda en ~/.config/yuy-chat/conversations/
+```
+
+### Cargar Conversación Anterior
+
+```
+Ctrl+C → 4 (Load Conversation)
+↑/↓ para navegar
+Enter para cargar
+```
+
+### Prompt Engineering
+
+**Para mejores respuestas, sé específico:**
+
+❌ Malo:
+```
+You: Explica Rust
+```
+
+✅ Bueno:
+```
+You: Explica el sistema de ownership en Rust con un ejemplo simple de borrowing. Quiero entender por qué evita memory leaks.
+```
+
+### Conversaciones Multi-paso
+
+```
+You: Vamos a diseñar una API REST
+
+Yuuki: Claro, ¿qué tipo de API?
+
+You: Para gestionar tareas tipo TODO
+
+Yuuki: Perfecto, estos son los endpoints...
+```
+
+### Usar Presets Dinámicamente
+
+- **Creative preset**: "Escribe un cuento de terror"
+- **Precise preset**: "¿Cuál es la complejidad de quicksort?"
+- **Balanced preset**: "Explícame cómo funciona Git"
+
+---
+
+## 🔧 Troubleshooting
+
+### Error: "No models found"
+
+**Solución:**
+```bash
+# Verifica que tienes modelos
+ls ~/.yuuki/models/
+
+# Si está vacío, descarga uno
+yuy download Yuuki-best
+
+# O especifica otro directorio
+yuy-chat --models-dir /path/to/models
+```
+
+### Error: "llama.cpp binary not found"
+
+**Solución:**
+```bash
+# Termux
+pkg install llama-cpp
+
+# macOS
+brew install llama.cpp
+
+# Linux - verifica que está en PATH
+which llama-cli
+# Si no, instala o agrega al PATH
+export PATH=$PATH:/path/to/llama-cpp
+```
+
+### Error: "Permission denied" (llamafile)
+
+**Solución:**
+```bash
+chmod +x ~/.yuuki/models/*.llamafile
+```
+
+### Chat no responde / se congela
+
+**Diagnóstico:**
+1. Verifica que llama.cpp funciona:
+```bash
+llama-cli -m ~/.yuuki/models/Yuuki-best.gguf -p "Hola"
+```
+
+2. Revisa logs:
+```bash
+RUST_LOG=debug yuy-chat
+```
+
+3. Reduce context size si es falta de RAM
+
+### Respuestas muy lentas
+
+**Causas comunes:**
+- Modelo muy grande para tu RAM
+- Cuantización muy alta (F32, Q8)
+- Sin aceleración GPU
+
+**Solución:**
+```bash
+# Descarga versión cuantizada más pequeña
+yuy download Yuuki-best --quant q4_0
+
+# Verifica RAM disponible
+free -h  # Linux
+top      # macOS/Linux
+```
+
+### No puedo escribir mensajes largos
+
+El input box tiene límite visual pero **no de contenido**:
+- Usa `Shift+Enter` para multi-línea
+- Scroll automático después de 5 líneas
+- O escribe en editor externo y pega
+
+### HuggingFace API no funciona
+
+**Verifica:**
+```bash
+# Test manual
+curl https://api-inference.huggingface.co/models/OpceanAI/Yuuki-best \
+  -H "Authorization: Bearer YOUR_TOKEN" \
+  -d '{"inputs": "test"}'
+```
+
+**Problemas comunes:**
+- Token expirado → Genera nuevo
+- Rate limit → Espera o upgrade plan
+- Modelo privado → Verifica permisos
+
+---
+
+## 📊 Performance Tips
+
+### Termux/Móvil
+
+```bash
+# Usa modelos pequeños
+yuy download Yuuki-best --quant q4_0
+
+# Preset Balanced o Precise
+# Creative es más lento
+```
+
+### Desktop High-end
+
+```bash
+# Usa Q8 o F32 para mejor calidad
+yuy download Yuuki-best --quant q8_0
+
+# Habilita GPU en llama.cpp
+llama-cli -m model.gguf -ngl 32  # 32 layers en GPU
+```
+
+---
+
+## 🎓 Casos de Uso
+
+### 1. Coding Assistant
+
+```
+Preset: Precise
+
+You: Cómo implemento un servidor HTTP en Rust?
+You: Muestra ejemplo con tokio
+You: Agrega manejo de errores
+You: Ahora agrega logging
+```
+
+### 2. Creative Writing
+
+```
+Preset: Creative
+
+You: Escribe el inicio de una novela de ciencia ficción ambientada en Marte en el año 2157
+You: Continúa describiendo al protagonista
+You: ¿Qué conflicto enfrenta?
+```
+
+### 3. Learning/Study
+
+```
+Preset: Balanced
+
+You: Explícame la diferencia entre mutex y semaphore
+You: Dame un ejemplo de cuándo usar cada uno
+You: ¿Qué pasa si no uso sincronización?
+```
+
+---
+
+## 🚀 Workflow Recomendado
+
+### Developer
+
+```bash
+# Mañana: Coding
+yuy-chat  # Preset: Precise
+> Ayuda con bugs, arquitectura, código
+
+# Tarde: Docs
+yuy-chat  # Preset: Balanced
+> Escribir documentación, READMEs
+
+# Noche: Ideas
+yuy-chat  # Preset: Creative
+> Brainstorming features
+```
+
+### Writer
+
+```bash
+yuy-chat  # Preset: Creative
+> Generar ideas
+> Escribir borradores
+> Feedback de historias
+```
+
+### Estudiante
+
+```bash
+yuy-chat  # Preset: Balanced
+> Explicaciones de conceptos
+> Resolver dudas
+> Preparar exámenes
+```
+
+---
+
+**¿Preguntas? Abre un issue en GitHub!**
+
+🌸 Hecho con amor por el equipo Yuuki
diff --git a/yuy-chat-complete/src/app.rs b/yuy-chat/src/app.rs
similarity index 100%
rename from yuy-chat-complete/src/app.rs
rename to yuy-chat/src/app.rs
diff --git a/yuy-chat-complete/src/config.rs b/yuy-chat/src/config.rs
similarity index 91%
rename from yuy-chat-complete/src/config.rs
rename to yuy-chat/src/config.rs
index a121669..ac2c82e 100644
--- a/yuy-chat-complete/src/config.rs
+++ b/yuy-chat/src/config.rs
@@ -110,3 +110,9 @@ impl Config {
         Ok(config_dir)
     }
 }
+
+// Yuuki constants
+pub const HF_ORG: &str = "OpceanAI";
+pub const OLLAMA_ORG: &str = "aguitachan3";
+pub const YUUKI_API: &str = "https://huggingface.co/spaces/OpceanAI/Yuuki-api";
+pub const AVAILABLE_QUANTS: &[&str] = &["q4_0", "q4_k_m", "q5_k_m", "q8_0", "f32"];
diff --git a/yuy-chat-complete/src/conversation.rs b/yuy-chat/src/conversation.rs
similarity index 100%
rename from yuy-chat-complete/src/conversation.rs
rename to yuy-chat/src/conversation.rs
diff --git a/yuy-chat-complete/src/main.rs b/yuy-chat/src/main.rs
similarity index 100%
rename from yuy-chat-complete/src/main.rs
rename to yuy-chat/src/main.rs
diff --git a/yuy-chat-complete/src/models/hf_api.rs b/yuy-chat/src/models/hf_api.rs
similarity index 100%
rename from yuy-chat-complete/src/models/hf_api.rs
rename to yuy-chat/src/models/hf_api.rs
diff --git a/yuy-chat-complete/src/models/mod.rs b/yuy-chat/src/models/mod.rs
similarity index 100%
rename from yuy-chat-complete/src/models/mod.rs
rename to yuy-chat/src/models/mod.rs
diff --git a/yuy-chat-complete/src/models/runtime.rs b/yuy-chat/src/models/runtime.rs
similarity index 76%
rename from yuy-chat-complete/src/models/runtime.rs
rename to yuy-chat/src/models/runtime.rs
index 8fbae15..3073b44 100644
--- a/yuy-chat-complete/src/models/runtime.rs
+++ b/yuy-chat/src/models/runtime.rs
@@ -1,5 +1,5 @@
 use super::{Model, ModelFormat, ModelSource};
-use crate::config::Preset;
+use crate::config::{Preset, YUUKI_API};
 use anyhow::{Context, Result};
 use std::process::Stdio;
 use tokio::io::{AsyncBufReadExt, BufReader};
@@ -109,19 +109,44 @@ impl ModelRuntime {
     }
 
     async fn generate_hf(&mut self, prompt: &str) -> Result<()> {
-        // Placeholder for HuggingFace API call
+        // Use Yuuki API
         let (tx, rx) = mpsc::channel(100);
         self.response_rx = Some(rx);
 
         let prompt_owned = prompt.to_string();
+        let api_url = YUUKI_API.to_string();
+        let temp = self.preset.temperature();
+        let top_p = self.preset.top_p();
         
         tokio::spawn(async move {
-            // Simulated streaming response
-            let response = format!("Response to: {}", prompt_owned);
-            for word in response.split_whitespace() {
-                let _ = tx.send(format!("{} ", word)).await;
-                tokio::time::sleep(tokio::time::Duration::from_millis(100)).await;
+            // Call Yuuki API
+            let client = reqwest::Client::new();
+            let response = client
+                .post(&api_url)
+                .json(&serde_json::json!({
+                    "prompt": prompt_owned,
+                    "temperature": temp,
+                    "top_p": top_p,
+                    "max_tokens": 512
+                }))
+                .send()
+                .await;
+
+            match response {
+                Ok(resp) => {
+                    if let Ok(text) = resp.text().await {
+                        // Stream response word by word
+                        for word in text.split_whitespace() {
+                            let _ = tx.send(format!("{} ", word)).await;
+                            tokio::time::sleep(tokio::time::Duration::from_millis(50)).await;
+                        }
+                    }
+                }
+                Err(_) => {
+                    let _ = tx.send("Error: Could not connect to Yuuki API".to_string()).await;
+                }
             }
+            
             let _ = tx.send("[DONE]".to_string()).await;
         });
 
diff --git a/yuy-chat-complete/src/models/scanner.rs b/yuy-chat/src/models/scanner.rs
similarity index 100%
rename from yuy-chat-complete/src/models/scanner.rs
rename to yuy-chat/src/models/scanner.rs
diff --git a/yuy-chat-complete/src/ui/chat.rs b/yuy-chat/src/ui/chat.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/chat.rs
rename to yuy-chat/src/ui/chat.rs
diff --git a/yuy-chat-complete/src/ui/conversations.rs b/yuy-chat/src/ui/conversations.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/conversations.rs
rename to yuy-chat/src/ui/conversations.rs
diff --git a/yuy-chat-complete/src/ui/menu.rs b/yuy-chat/src/ui/menu.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/menu.rs
rename to yuy-chat/src/ui/menu.rs
diff --git a/yuy-chat-complete/src/ui/mod.rs b/yuy-chat/src/ui/mod.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/mod.rs
rename to yuy-chat/src/ui/mod.rs
diff --git a/yuy-chat-complete/src/ui/selector.rs b/yuy-chat/src/ui/selector.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/selector.rs
rename to yuy-chat/src/ui/selector.rs
diff --git a/yuy-chat-complete/src/ui/settings.rs b/yuy-chat/src/ui/settings.rs
similarity index 100%
rename from yuy-chat-complete/src/ui/settings.rs
rename to yuy-chat/src/ui/settings.rs