SwiftChat — A Cross-platform AI Assistant

🚀 Your Personal AI Workspace — Chat, Create Apps, and More

中文

SwiftChat is a fast and responsive AI assistant developed with React Native and powered by Amazon Bedrock, with compatibility extending to other model providers such as Ollama, DeepSeek, OpenAI and OpenAI Compatible. With its minimalist design philosophy and robust privacy protection, it delivers real-time streaming conversations, AI image generation, instant web app creation and voice conversation capabilities across Android, iOS, and macOS platforms.

What's New 🔥

• 🚀 Support creating multiple apps in background with Live Activity on iOS and notifications on Android (From v2.8.0).
• 🚀 Upgrade to React Native 0.83 with New Architecture enabled (From v2.8.0).
• 🚀 Support create or edit instant web apps with one prompt (From v2.7.0).
• 🚀 Support Web Search for real-time information retrieval (From v2.7.0).
• 🚀 Update SwiftChat Server with API Gateway + Lambda deployment supporting 15-minute streaming output (From v2.7.0).
• 🚀 Support Image Gallery for browsing and managing generated images (From v2.7.0).

Create App: Generate, Edit, Share and Preview Instant Web Apps

App Examples: 2048 Game, Gomoku, Tetris and News Reader

Web Search & Mermaid: Real-time Information Retrieval and Streaming Chart Rendering

Note: Tavily is recommended for best results. Google Search requires manual verification on first use. Baidu and Bing are currently in beta and may return inaccurate results.

📱 Quick Download

• Download for Android
• Download for macOS
• For iOS: Currently available through local build with Xcode

Getting Started with Amazon Bedrock

Configuration

You can choose one of the following two methods for configuration

🔧 Configure Bedrock API Key (Click to expand)

1. Click Amazon Bedrock Console to create a long-term API key.

2. Copy and paste the API key to the (Amazon Bedrock -> Bedrock API Key) under SwiftChat Settings page.

3. The App will automatically get the latest model list based on the region you currently selected. If multiple models appear in the list, it means the configuration is successful.

🔧 Configure SwiftChat Server (Click to expand)

Architecture

We use API Gateway combined with AWS Lambda to enable streaming responses for up to 15 minutes. All requests are authenticated via API Gateway's API Key validation before being forwarded to Lambda, ensuring secure access to backend services.

Step 1: One-command deploy

Prerequisites: AWS CLI configured with credentials (aws configure).

Deploy with one command:

curl -fsSL https://raw.githubusercontent.com/aws-samples/swift-chat/main/server/install.sh | bash

By default this deploys to the account of your current AWS profile, in the region from $AWS_REGION or your aws configure default (fallback: us-east-1), with stack name SwiftChat. Wait about 3-4 minutes to see your API URL, API Key obtain link, and a scannable QR code for auto-configuring the app.

To target a different account, region, or stack:

./install.sh --region us-west-2 --stack MySwiftChat --profile myprofile

Flag	Purpose	Default
--region	AWS region to deploy into	$AWS_REGION or us-east-1
--stack	CloudFormation stack name	SwiftChat
--profile	AWS CLI profile to use	current default profile

Step 2: Open the App and setup with API URL and API Key

1. Launch the App, open the drawer menu, and tap Settings.
2. Under Amazon Bedrock → SwiftChat Server, tap the scan icon next to API URL to scan the QR (iOS/Android) — or paste API URL and API Key manually.
3. Select your Region, then tap the top right ✓ to save.

Congratulations 🎉 Your SwiftChat App is ready to use!

Getting Started with Other Model Providers

Ollama

🔧 Configure Ollama (Click to expand)

1. Navigate to the Settings Page and select the Ollama tab.
2. Enter your Ollama Server URL. For example: bash http://localhost:11434

3. Enter your Ollama Server API Key (Optional).

4. Once the correct Server URL is entered, you can select your desired Ollama models from the Chat Model dropdown list.

DeepSeek

🔧 Configure DeepSeek (Click to expand)

1. Go to the Settings Page and select the DeepSeek tab.
2. Input your DeepSeek API Key.
3. Choose DeepSeek models from the Chat Model dropdown list. Currently, the following DeepSeek models are supported:
   • DeepSeek-Chat
   • DeepSeek-Reasoner

OpenAI

🔧 Configure OpenAI (Click to expand)

1. Navigate to the Settings Page and select the OpenAI tab.
2. Enter your OpenAI API Key.
3. Select OpenAI models from the Chat Model dropdown list. The following OpenAI models are currently supported:
   • GPT-5.4
   • GPT-5.4 Pro
   • GPT-5.4 mini
   • GPT-5.4 nano
   • GPT-5.3 Chat
   • GPT-5.2
   • GPT-5.2 Pro
   • GPT-5.1
   • GPT-5
   • GPT-5 Pro
   • GPT-5 mini
   • GPT-5 nano
   • GPT-4.1
   • GPT-4.1 mini
   • GPT-4.1 nano
   • GPT-4o
   • GPT-4o mini

Additionally, if you have deployed and configured the SwiftChat Server, you can enable the Use Proxy option to forward your requests.

OpenAI Compatible

🔧 Configure OpenAI Compatible models (Click to expand)

1. Navigate to the Settings Page and select the OpenAI tab.
2. Under OpenAI Compatible, enter the following information:
   • Base URL of your model provider
   • API Key of your model provider
   • Model ID of the models you want to use (separate multiple models with commas)
3. Select one of your models from the Chat Model dropdown list.
4. Click the plus button on the right to add another OpenAI-compatible model provider. You can add up to 10 OpenAI-compatible model providers.

Key Features

• Real-time streaming chat with AI
• Instant web app creation, editing and sharing
• Background app creation with progress tracking
• Web search for real-time information retrieval
• Rich Markdown Support: Tables, Code Blocks, LaTeX, Mermaid Chart and More
• AI image generation with progress
• Multimodal support (images, videos & documents)
• Conversation history list view and management
• Cross-platform support (Android, iOS, macOS)
• Tablet-optimized for iPad and Android tablets
• Fast launch and responsive performance
• Multiple AI models supported (Amazon Bedrock, Ollama, DeepSeek, OpenAI and OpenAI Compatible Models)
• Fully Customizable System Prompt Assistant

Feature Showcase

Comprehensive Multimodal Analysis: Text, Image, Document and Video

Creative Image Suite: Generation, Virtual try-on, Background Removal and Image Gallery with Nova Canvas

System Prompt Assistant: Useful Preset System Prompts with Full Management Capabilities (Add/Edit/Sort/Delete)

Rich Markdown Support: Paragraph, Code Blocks, Tables, LaTeX, Mermaid and More

We redesigned the UI with optimized font sizes and line spacing for a more elegant and clean presentation. All of these features are also seamlessly displayed on Android and macOS with native UI

Dark Mode

Note: Some animated images have been sped up for demonstration. If you experience lag, please view on Chrome, Firefox, or Edge browser on your computer.

Amazon Nova Series Features

Easy to use Virtual try-on by Nova Canvas

1. Support automatic setting of the main image, default is the previously used main image.
2. Support uploading or taking the second image and sending it directly without any prompt words.
3. Support automatically recognizes clothes, pants, shoes and tries them on

Amazon Nova Sonic Speech to Speech Model

1. Built-in spoken language practice for words and sentences, as well as storytelling scenarios. You can also add Custom System Prompts for voice chatting in different scenarios.
2. Support Barge In by default, Also you can disable in system prompt.
3. Support selecting voices in the settings page, including American/British English, Spanish and options for male and female voices.
4. Support Echo Cancellation, You can talk directly to the device without wearing headphones.
5. Support Voice Waveform to display volume level.

Learn Sentences

https://github.com/user-attachments/assets/ebf21b12-9c93-4d2e-a109-1d6484019838

Telling Story on Mac (With barge in feature)

https://github.com/user-attachments/assets/c70fc2b4-8960-4a5e-b4f8-420fcd5eafd4

Note: Amazon Nova Sonic currently only available with SwiftChat server.

Other Features

• Record 30-second videos directly on Android and iOS for Nova analysis
• Upload large videos (1080p/4K) beyond 8MB with auto compression
• Support using default template to make Nova Canvas generate images, remove backgrounds, and create images in similar styles.

Detailed Features

Quick Access Tools: Code & Content Copy, Selection Mode, Model Switch, Regenerate, Scroll Controls and Token Counter

We feature streamlined chat History, Settings pages, and intuitive Usage statistics:

Message Handling

• [x] Text copy support:
  • Copy button at the bottom of messages, or directly click the model name or user title section.
  • Copy button in code blocks
  • Copy button in reasoning blocks
  • Direct Select and copy code on macOS (double-click or long click on iOS)
  • Long press text to copy the entire sentence (Right-click on macOS)
• [x] Text selection mode by click selection button.
• [x] Message timeline view in history
• [x] Delete messages through long press in history
• [x] Click to preview for documents videos and images
• [x] Support for collapsing and expanding the reasoning section and remembering the most recent state

Image Features

• [x] Support image generation with Chinese prompts (Make sure Amazon Nova Lite is enabled in your selected region)
• [x] Long press images to save or share
• [x] Automatic image compression to improve response speed
• [x] Image Gallery for browsing and managing all generated images

User Experience

• [x] Haptic feedback for Android and iOS (can be disabled in Settings)
• [x] Support landscape mode on Android/iOS devices
• [x] Double tap title bar to scroll to top
• [x] Click bottom arrow to view the latest messages
• [x] Display system prompt and model switch icon again by clicking on the chat title
• [x] View current session token usage by tapping twice Chat title
• [x] Check detailed token usage and image generation count in Settings
• [x] In-app upgrade notifications (Android & macO