Selene

Selene is a desktop AI agent app for coding, browser tasks and personal assistants. You can choose local or cloud processing separately for chat, embeddings, voice and image generation.

• Coding and UI design: Agents can work with Git changes and pull requests, run checks after edits, and generate React or HTML components with sandbox previews and targeted revisions.
• Browser and screen tasks: Agents operate an embedded Chromium browser through Playwright, replay recorded browser actions and inspect your display through a vision sidecar.
• Personal assistants: Give agents access to synced folders, approve suggested memories and schedule tasks whose results stay in chat or go to WhatsApp, Telegram, Slack or Discord.
• Voice and images: Transcribe voice input and messaging voice notes, generate images locally or through cloud services, and use reference images or ComfyUI workflows; voice cloning uses ElevenLabs or Microsoft.
• Agent extensions: Add reusable skills, plugins and MCP tools, or let agents delegate tasks to parallel sub-agents; retrieval supplies relevant context and tools load when needed.

Selene is open source under the MIT license and runs on macOS and Windows. It supports fully offline use without API keys through local processing, including Ollama models, and can connect to self-hosted vLLM. Cloud choices include Claude, OpenAI and OpenRouter. Local processing keeps the relevant work on your device; cloud providers receive the work you send to them, and messaging integrations require connections to their services. The desktop download bundles local model support and the browser engine.