What is Gwen?

Gwen is Intelligence That Works. Most AI tools answer questions. Gwen does the work.

You describe an outcome in plain language — a website that needs to exist, a backlog of content, a research question, a CRM that needs cleaning up, a software change, an inbox that needs triaging — and Gwen scopes the job, does it across whatever tools and models the work requires, shows you the result, and asks before anything leaves your control. Think of Gwen less like a chatbot and more like a capable new hire who already knows your context, never forgets a decision, and works in two modes: Enterprise Intelligence for your work, and Personal Intelligence for your life.

There is a difference between getting an answer and getting the thing done. A chat tool can tell you how to build a landing page. It can draft an email if you paste in the context every time. It can suggest a plan. But the moment the job involves more than one step, more than one tool, or more than one session, you become the operator: copying outputs between apps, re-explaining who you are, stitching the pieces together yourself.

Gwen is built to remove that operator role. The deliverable is the point — a live preview you can click, a content campaign ready to review, a research brief with sources, a pull request, a cleaned-up contact list — not a transcript you still have to act on.

You do not have to know which model, which framework, or which integration the job needs. That is Gwen's job to figure out.

You say what you want to be true — "I need a booking site for my home-inspection business," "turn last quarter's calls into a follow-up sequence," "find out who the serious competitors are and where they're weak" — and Gwen translates that into a scoped plan: the work area, the likely deliverable, the tools and connections required, the approvals that will be needed, and the first executable step. You approve the shape of the work, and Gwen executes it.

Enterprise Intelligence is Gwen's understanding of how your business actually runs: your goals, your customers and leads, your documents and decisions, your tone of voice, your work areas, the tools you use, and the history of what has already been done.

Because that context lives with Gwen instead of in your head, work that used to require a long briefing becomes a short request. Sales, marketing, content, research, operations, support, software, and the day-to-day coordination between them can all draw on the same shared understanding of the company.

Personal Intelligence is the same idea, pointed at the individual. Gwen can carry how you prefer to work and decide: your priorities, your communication style, the people and projects that matter, the recurring tasks you would rather not think about, and the personal context that should not vanish between apps and sessions.

Work and life are rarely cleanly separated for the people doing the most important work. Gwen is designed to hold both, with a clear boundary between what belongs to a company workspace and what belongs to you personally.

Underneath, Gwen is not a single model. The hard part of any real job is choosing the right capability for each step and making the pieces work together — and Gwen does that for you.

Gwen draws on the leading AI engines: Claude, GPT-5, OpenAI Codex, Gemini, Llama, and Hermes, among others. Model access and routing run through Tokaroo (tokaroo.com), an intelligent router across hundreds of models from providers including Anthropic, DeepSeek, Google, Groq, Moonshot, and OpenAI that handles selection, fallback, and cost. The practical result: Gwen can use a frontier reasoning model to plan, a specialized coding engine to build, and a separate model to independently check the work — without you ever opening a model picker.

Gwen is meant to be asked for outcomes across the work that actually fills a week. A non-exhaustive sense of the range:

• Build and launch: websites, landing pages, app workspaces, client portals, and ongoing edits after launch.
• Grow: capture leads, keep a CRM, draft content and campaigns, and prepare nurture and follow-up.
• Research and analyze: competitor and market research, summaries, briefs, and decision support with sources.
• Software: scope changes, write and review code, prepare diffs and pull requests, and coordinate a repository.
• Operations and inbox: triage, draft replies, organize information, and turn recurring requests into workflows.
• Personal: keep track of priorities, decisions, people, and the context you do not want to re-explain.

A normal assistant starts more or less from your last message. Every session is a cold start; every tool is something you operate; every output is something you carry to the next step yourself.

Gwen is designed to carry context across missions, tools, memory, and workspaces, and to learn from feedback so the next request is easier than the last. The first time you ask for something it may need to ask questions and get set up. The tenth time, it already knows your brand, your audience, your constraints, and how you like the work delivered.

Doing the work does not mean acting behind your back. Gwen can freely read available context, scope work, draft, summarize, and prepare deliverables. But anything that touches the outside world — sending an email, publishing a page, spending money, pushing code, changing records, deleting things — asks for approval or follows a policy you set.

Every mission keeps a ledger: what Gwen did, what context it used, which tools it tried, what is waiting on you, and what it produced. You can inspect, correct, and steer at any point.

Because Gwen remembers approved context and learns from outcomes, the same workspace compounds. Repeated work becomes saved knowledge, reusable workflows, and playbooks. Corrections stick. Preferences carry forward.

Gwen also improves as a product: new engines, new integrations, and new capabilities arrive behind the same simple promise — you ask for the outcome, and the result keeps getting better and faster.

You start by telling Gwen what you want, in the chat. There is no setup project to complete first and no model to choose.

When a request should become real work, Gwen creates your workspace and runs an eligible first pass at no cost so you can see the quality before you commit. From there, a prepaid Work Budget funds continued execution — you pay for work done, not for seats. See How Gwen starts and Gwen Work Budget for the details.