Private AI Workspace for Mac | Elvean

Intel & Apple Silicon · macOS 13+

Rich Content

Itineraries with inline photos and formatted details

Sortable Tables

Structured data you can sort and explore inline

Built-in Tools

Web search, image search, MCP servers, and charts

Interactive Charts

Chart.js and TradingView charts rendered in conversations

Threaded Conversations

Branch into side threads without losing context

Agentic Research

Multi-step tool calling that fetches and presents findings

Photo Galleries

Image grids fetched and arranged automatically

Apple Maps

Interactive maps embedded in conversations with pinned locations

Your knowledge base grows as you work

Projects give every conversation shared context. Add persistent system instructions, attach files, and build a knowledge base that your model can search, read, and write — all mid-conversation. Turn on auto-distill and key facts get extracted into your KB automatically. No scripts, no terminal, no RAG pipeline.

• Projects with persistent system instructions and color coding
• Filesystem-based Knowledge Base — add files, create and edit pages
• Models search, read, and write KB pages during conversation
• Auto-distill: key facts extracted into KB after each response

Every model, one app

Stop switching between ChatGPT, Claude, and a dozen browser tabs. Elvean connects to every major AI provider and your local models in one native interface. Bring your own API keys, pay providers directly, no markup.

• Ollama, LM Studio, MLX, Claude, OpenAI, Gemini, Grok, Cerebras, DeepSeek
• 300+ models via OpenRouter — including any OpenAI-compatible API
• Switch models mid-conversation with @mentions
• API keys stored in macOS Keychain, never on our servers

Responses that come alive

AI responses shouldn't be walls of text. Elvean renders charts you can interact with, tables you can sort, maps you can explore, and image galleries you can browse — all inline, right in the conversation.

• Interactive Chart.js charts and Mermaid diagrams
• Sortable, filterable tables with structured data
• Embedded Apple Maps with pinned locations
• Photo galleries fetched and arranged automatically
• Real-time weather forecasts via Apple WeatherKit

Put AI to work on multi-step tasks

Move beyond one-off prompts. Elvean's agent mode handles work that needs planning, iteration, and follow-through in a single flow.

• Action-oriented workflows that plan the next step, not just answer
• Built-in tools: web search, Apple Calendar, weather, screenshot to AI
• MCP servers for files, databases, and any external API
• Multi-step execution with live progress and inline results

Give your AI a real environment

Your model doesn't just talk about code — it runs it. Each project can spin up a sandboxed Linux container with a full terminal. AI executes commands inside the sandbox and produces artifacts you can access from your project folder. Open the built-in terminal to interact with it directly.

• Isolated Linux container per project
• AI runs commands, installs packages, writes files
• Artifacts accessible from your project folder
• Built-in terminal panel (Cmd+Shift+T)

Conversations that stay organized

Long conversations get messy. Elvean brings Slack-style threads to AI — branch off into a side discussion without losing your main context. Fork any message to explore a different direction. Your thinking stays structured.

• Threaded conversations — branch without losing context
• Fork any message to explore alternate directions
• Folders, favorites, and full-text search across all conversations
• Cmd+F find within any conversation

Built for your Mac, not ported to it

Elvean is a native SwiftUI app — not an Electron wrapper, not a web view in a frame. It launches instantly, uses minimal memory, and respects every macOS convention you expect. Keyboard shortcuts, dark mode, and a system-wide prompt panel that works from any app.

• Native SwiftUI — optimized for Apple Silicon and Intel
• System-wide prompt panel: select text anywhere, hit a hotkey, get AI
• Prompt library with variable placeholders and Cmd+/ access
• Per-conversation settings: temperature, tokens, system prompts

🎙Voice input & text-to-speech

📎Image & PDF attachments

📝Edit any AI response in Markdown

📊Token usage tracking

📤Export to Markdown, HTML, PDF

⚡Metal-accelerated local inference