# Orkas > Open-source desktop app where a Commander agent coordinates specialist AI agents to produce research, decks, videos, code and apps. Orkas is an open-source, local-first desktop app for macOS and Windows that lets solo founders and small teams assign work to a group of AI agents through chat. In Orkas, a Commander agent splits goals among specialist agents that deliver reports, slides, videos, websites, code and recurring business tasks. **Best for:** Solo founders and small teams who want AI agents to handle research, content, design, development and routine operations without hiring extra staff. - Official site: https://orkas.ai - Listed in: [AI Agents & Automation](https://trendhub.net/ai-tools/category/ai-agents), [AI Code Assistant](https://trendhub.net/ai-tools/category/code-assistant), [Productivity & Team Work](https://trendhub.net/ai-tools/category/productivity-tools) - Last checked: 2026-11-03 (facts taken from orkas.ai) - Source page: https://trendhub.net/ai-tools/orkas-open-source-ai-agent-team-desktop ## Key facts - Platforms: Windows, macOS, Linux, API, MCP - Pricing model: Freemium - Free plan: Yes - API: Yes - Made by: Orkas ## What is Orkas? Orkas is a multi-agent AI desktop client, released under the MIT license, in which a Commander agent directs specialist agents such as a researcher, content writer, deck designer, video producer, UI designer, developer and SEO/GEO optimizer. Users describe a goal in chat and receive finished files like MP4 videos, HTML pages, PDF reports or PowerPoint decks. Repeated jobs can be scheduled as automations, and missed runs catch up after a computer wakes. Files, chats and memory are stored on the user's machine by default, with optional cloud sync. Orkas ships with managed official models and also accepts connections to providers such as OpenAI, Anthropic, Google Gemini, DeepSeek and OpenRouter, plus local coding agents including Claude Code and Codex. Teammates can share projects, and paid plans allow publishing compatible apps as web apps reachable by a link. ## Key features - **Commander and specialist agents**: A lead agent breaks a request into parts and hands them to role-specific agents for research, writing, images, video, audio, slides, office tasks, UI design, coding and SEO. - **Scheduled automations**: Tasks can run on a schedule, such as every Monday morning, and an auto-advance mode works through the project to-do list between check-ins. - **Persistent memory**: Preferences and facts from chats, such as tone or product details, are saved per project and applied to later tasks; users can review, edit or delete them. - **Custom agent creation**: Describing a missing role in plain language produces a new agent with a written workflow that can be edited in the AI Team section. - **Model and agent choice**: Official managed models work immediately, or users connect their own provider via OAuth or API key and bring local agents like Claude Code, Codex, OpenCode, OpenClaw and Hermes. - **Web app publishing and shared projects**: Compatible apps such as quote tools can be published to a public link with view statistics, and teammates can join shared projects with common files, goals and to-dos. ## How Orkas works 1. **Install the desktop app**: Download Orkas for macOS or Windows and sign in to claim the starting credits, or connect your own model provider. 2. **Give a goal in chat**: Describe the job you need, such as a launch kit or a client deck, and the Commander assigns it to specialist agents. 3. **Receive finished files**: The agents return deliverables like videos, documents, slides or code inside the project. 4. **Turn it into a recurring task**: Schedule the job to repeat, for example weekly, so it keeps running between your check-ins. ## Use cases - **Product launch content**: An online store asks for a launch kit containing a short captioned video, product page copy and a week of social posts in its brand voice. - **Weekly store review**: A small e-commerce shop gets a Monday brief on sales, orders and reviews, plus drafted review replies and follow-up to-dos. - **Agency client reporting**: A marketing agency combines ad account numbers into a client update deck in its template and receives briefs before client calls. - **Indie software development**: A solo developer has agents design an interface, add a search feature and fix a login bug with tests, then summarize changes in a changelog. - **Customer-facing tools**: A business builds a wholesale quote calculator and publishes it as a web app that customers open from a shared link. ## Things to know - **Published app limits**: Web apps published from Orkas cannot directly access desktop AI, local files or connectors, and publishing requires an eligible paid plan. - **Credit rules**: Credits only meter the Orkas-managed services; calls through your own provider key are billed by that provider. Monthly credits do not roll over and are spent before lifetime credits. - **Welcome credits**: The 500 bonus credits for new users are granted only when signing in through the desktop app, not through web login. - **Platforms and languages**: Official builds cover macOS and Windows; Linux is supported by building from the open-source repository. The interface has 18 language options. ## FAQs ### What is Orkas? Orkas is an open-source, local-first multi-agent AI desktop app for macOS and Windows. In Orkas, a Commander agent coordinates specialist agents to complete complex work. ### Is Orkas free? Orkas has a free desktop app that includes 500 bonus credits, built-in models, your own API keys and limited project collaboration. Paid Orkas plans add monthly credits, multi-device sync, more cloud storage and web app publishing. ### Is Orkas open source? Yes, the Orkas desktop client is open source under the MIT license and hosted on GitHub. Users can read the code, build it themselves and check what the app sends over the network. ### Do I need my own model API key to use Orkas? No, Orkas provides managed official models that work out of the box. Orkas also lets you connect providers such as OpenAI, Claude, Gemini, DeepSeek, Kimi and GLM via OAuth or API key. ### Does Orkas upload my files? By default, Orkas keeps configs, history, knowledge bases and generated files on your machine. Orkas services are used only for managed models and optional cloud features you select. ### Does Orkas have an API? Yes, Orkas offers an Open API so membership credits can be used in external products with an API key. ### Which platforms does Orkas support? Orkas offers official builds for macOS and Windows. The open-source Orkas repository also supports Linux if you build it yourself. ### Can I self-host Orkas? Yes, the open-source Orkas edition can run on your own machine without an Orkas account using your own provider keys. Orkas is a desktop client rather than a server, so self-hosting does not mean deploying a team server. --- Source: https://trendhub.net/ai-tools/orkas-open-source-ai-agent-team-desktop (trendhub)