**Wingman โ AI Chat for Power Users**
Take full control of your AI conversations. Wingman is a free, open-source chat app that connects directly to the world's best AI models using your own API keys โ no subscriptions, no lock-in, no data sent anywhere except straight to the AI provider of your choice.
**๐ Four AI Providers in One App**
Connect to Google Gemini, OpenAI (ChatGPT & DALL-E), Anthropic Claude, and DeepSeek from a single, unified interface. Switch between models mid-conversation and pick the right tool for every task.
**๐ผ๏ธ Multimodal โ Not Just Text**
Attach images, PDFs, and documents to your messages. Generate images with DALL-E directly in the chat. Save generated images to your device with one tap.
**๐ Real-Time Streaming**
Watch responses stream in token by token โ just like the official apps โ with a stop button to interrupt at any time.
**๐ก๏ธ Temperature Presets**
One tap to switch between six AI creativity levels: Coding, Strict, Analysis, Conversation, Translation, and Creative.
**๐ฌ Full Chat History**
All your conversations are saved locally on your device. Browse, resume, or delete sessions from the sidebar. Crash-safe โ your history survives even if the app closes unexpectedly mid-response.
**๐จ Beautiful, Adaptive UI**
A clean, modern interface that follows your system's dark or light theme. Markdown rendering with syntax-highlighted code blocks in assistant replies.
**๐ Privacy First**
Your API keys never leave your device except to talk directly to the AI provider. No analytics, no tracking, no Wingman servers. The full source code is available on GitHub.
**โ
Free & Open Source**
Wingman is completely free โ no ads, no in-app purchases, no premium tier. Built with .NET MAUI and published under the BSD 3-Clause License.
---
**What you need:**
An API key from at least one supported provider (Google AI Studio, OpenAI, Anthropic, or DeepSeek). All providers offer free tiers or pay-as-you-go pricing โ you only pay for what you use.
---
**Known Limitations:**
- File attachments from earlier in a conversation are not re-sent to the AI in follow-up messages (current message only)
- Voice input UI is not available in this release
- iOS not supported yet