Gwen FAQ

Gwen is Intelligence That Works: Enterprise Intelligence for your business and Personal Intelligence for your life. Where most AI tools answer questions, Gwen does the work — it understands your context, scopes a job, produces real deliverables using the right engines and tools, and asks before it takes any action in the outside world.

A chat assistant gives you answers and largely starts fresh each session, leaving you to operate the tools and carry outputs from step to step. Gwen is built to do the work end to end and to remember you across sessions, tools, and missions.

It carries context, coordinates multiple engines, produces reviewable artifacts, and improves from feedback — so the tenth request takes far less briefing than the first.

Ask for outcomes, not features. Gwen can build and launch websites and app workspaces, run a growth loop (lead capture, CRM, content, nurture), do research and analysis, take on software work, and handle operations and inbox work — across both your business and your personal life. See "What Gwen can do" for the range.

Gwen is not a single model. It draws on leading engines — Claude, GPT-5, OpenAI Codex, Gemini, Llama, and Hermes among others — and routes across hundreds of models from providers including Anthropic, DeepSeek, Google, Groq, Moonshot, and OpenAI through Tokaroo (tokaroo.com) for selection, fallback, and cost.

You never choose a model. Gwen picks the right engine for each step, and can use one model to build and a different one to independently check the work.

Yes. Gwen is designed to build a living, connected memory of useful context — goals, preferences, decisions, documents, people, projects, and outcomes — so it stops asking the same questions and work compounds over time.

Memory stays under your control: you can inspect it, correct it, approve what becomes durable, and remove what should not persist.

Yes, when a tool is connected to your workspace and a job needs it. Gwen can work through email, a CRM, documents and drives, calendars, browsers, repositories, and other business systems.

Connections are scoped to your workspace and used only for what you authorized. Gwen tells you when a job needs access you have not granted rather than reaching for it silently.

Yes — and you can see it for free first. For an eligible first build, Gwen produces a hosted preview you can open and judge before paying anything. Continued edits, publishing, and launch are funded by a Work Budget and gated by approval where they touch the live world.

Gwen can create and manage durable project workspaces, write and edit code, run builds and tests, and prepare diffs and pull requests for review. Repository pushes, pull requests, and deploys require your approval and the right connected workspace.

Gwen is priced for work done, not seats filled. You fund a prepaid Work Budget that is consumed as approved work runs; unused credit expires after 12 months of workspace inactivity, and auto top-up is authorized explicitly at checkout and can be paused later.

A new workspace gets a free first pass so you can see real output before committing. For current amounts, see the pricing page.

No. Gwen reads, plans, drafts, and prepares freely, but it asks before anything irreversible or outward-facing — sends, publishing, spend, deployments, record changes, deletions, and other sensitive actions — either prompting you or following a policy you set.

Your data is yours and is used to do your work, not to become someone else's product. Each workspace is isolated from every other, company knowledge is kept separate from personal memory, and you can inspect, correct, and remove what Gwen knows. See "Approvals, control, and your data."

Both. Enterprise Intelligence handles company work; Personal Intelligence handles your individual context. They share one intelligence layer with a clear boundary between company and personal data, so the same Gwen can help across your whole world without mixing the two.