The AI Chat Interface You've Been Waiting For

Provider & MCP Marketplace

One-click setup for 15+ AI providers and 20+ MCP servers. Same look, same flow — no JSON editing required.

Multi-model Comparison

Run the same prompt against 2–4 models side-by-side simultaneously. See responses, latency, and token counts at a glance.

File & Image Attachments

Attach images (vision), PDFs, and text files to any message. Vision content is passed natively to OpenAI, Anthropic, and Gemini.

MCP Tool Servers

Extend the AI with Model Context Protocol servers over HTTP-SSE or stdio. One-click install from the built-in marketplace.

Markdown + Code

Full markdown rendering with syntax-highlighted code blocks via highlight.js.

Conversation History

Persistent conversation storage. All your chats, always available, stored locally.

Privacy First

API keys stored encrypted on-device. No telemetry, no data collection by default.

Cross-Platform

Native builds for macOS (Apple Silicon), Windows, and Linux (deb/rpm).

Open Source

AGPL-3.0 open source. Fully auditable, forkable, and self-hostable. Commercial licenses available.

Best-in-class general reasoning, coding, and multimodal model support.

Supported models

GPT-5.1, GPT-4.1, GPT-4.1 mini, GPT-4o, o3, o4-mini

Excellent long-context and instruction-following workflows for deep technical tasks.

Supported models

Claude Opus 4.7, Claude Sonnet 4.6, Claude 3.7 Sonnet, Claude 3.5 Haiku

Google’s frontier models with a generous free tier and top-tier multimodal capabilities.

Supported models

Gemini 2.5 Pro, Gemini 2.0 Flash, Gemini 2.0 Flash Lite

Run local/offline models with full control over privacy and hardware usage.

Supported models

Llama 3.3, Qwen 2.5, DeepSeek R1, Mistral, Phi-4, Gemma 2 — and more