What it does
Orkas is a local-first multi-agent desktop app. You describe a goal in one chat, and its Commander plans the work, handles the general parts itself and coordinates specialist agents in parallel or in sequence. Nine specialist agents ship with the app, covering research, writing, slide decks, engineering, office files, video, images, UI design and SEO audits, and an agent marketplace lists 30. Orkas can also run installed CLI agents as local sessions: Claude Code, Codex and OpenCode, plus OpenClaw and Hermes-Agent. Conversations, files, API keys, knowledge bases and custom agents stay on your disk, and model calls go straight from your machine to the provider you choose. You can mix providers across agents, including Claude, OpenAI, Gemini, DeepSeek, Kimi, GLM, Qwen, MiniMax and Doubao, or point an agent at a local OpenAI-compatible or Anthropic-compatible endpoint. The app is MIT licensed. Packaged installers cover macOS and Windows 10 or later, and Linux x64 and arm64 run from source. The packaged desktop app asks you to sign in on first launch, while the source build has no account layer.
Capabilities
- Commander agent that plans a goal and coordinates specialist agents in parallel or in sequence
- Nine built-in specialist agents, with 30 in the agent marketplace
- Drives Claude Code, Codex, OpenCode, OpenClaw and Hermes-Agent as local sessions
- Bring your own model keys and mix providers across agents
- Conversations, files, keys and knowledge bases stay on your own disk
- Each agent keeps its own private skills and memory and improves through reflection
When to use it
- You want one desktop chat that splits a goal across research, writing, slides and engineering agents
- You want to run several coding CLIs side by side from a single interface
- You want your data and API keys kept on your own machine rather than in a vendor cloud
- You already have model API keys or a local model endpoint and want to pick the model per agent
