Floutwork

Floutwork is a desktop application for Mac and Windows that tries to collapse the scattered workday — AI chats, browser tabs, notes, tasks, calendar, inbox — into one governed workspace. It began life as an AI productivity browser aimed at individuals; at time of writing it describes itself as an "AI Work OS, Built for Teams," and the pivot shows in everything from the pricing page to the admin console. The pitch to a company is consolidation: instead of every employee holding separate ChatGPT, Claude, Gemini, Perplexity and Grok subscriptions plus a task app and a notes app, one Floutwork seat covers the models and the workflow layers around them, with an administrator deciding which models people may use.

What it actually does

The AI layer is the headline: one chat surface with several frontier model families behind it — OpenAI GPT, Anthropic Claude, Google Gemini, with Perplexity and Grok named in the vendor's savings calculator. Employees switch models without switching tools, and custom AI assistants can be built for recurring workflows (drafting, research, analysis). Centralized governance sits above this: admins control which models and assistants are available and set usage policies. A "personal memory engine" is the vendor's term for the context layer — it learns from notes, tasks, chats and bookmarks and feeds relevant context back into assistant responses.

The workspace layer is what makes Floutwork a place rather than a chat wrapper. It runs as its own desktop app (Apple Silicon and Intel Macs, Windows) with a browser whose tabs can be grouped by application or project, split views so a document and its task list sit side by side, and named workspaces per team, department or project — each with its own tools, policies and AI capabilities. New employees can be onboarded into a preconfigured workspace. Tasks and notes can be created from any page and synced to a calendar; a "smart inbox cleaner" batches email for bulk action; productivity metrics score work patterns. Google and Outlook connections are supported, with login details stored on the local machine. Hosting is on AWS in the US, and the vendor states it does not train models on customer data.

Pricing at time of writing is simple and team-shaped. The pricing page shows Pro at $29 per user per month excluding VAT — annual billing, with monthly billing costing about 20% more — covering up to 1,000 users, all features, and SSO/SCIM/SAML. Enterprise is quoted for over 1,000 users with dedicated deployment options. A 7-day trial includes $5 of AI credit and requires no card. The page also carries a savings calculator claiming a 25-person team paying for ChatGPT Team, Gemini Business, Claude Team, Perplexity Pro and Grok would drop from $3,350/month to $725/month — a claimed 78% saving. That is a vendor calculation with vendor-chosen inputs; I treat it as a hypothesis to test against your own usage, not a fact.

My testing notes

First observation: the consolidation story holds together better than I expected at the workspace level. In my testing, the value was not any single feature but the removal of boundaries between them — referencing a task in a chat, pinning a note beside the tab it came from, switching models mid-thread without re-explaining context. The workspace concept works: I set up one per project, and switching between them genuinely reset the context of my day. For someone whose job is juggling a dozen web tools, that is a real reduction in friction, and the tab-grouping browser is the piece doing the quiet work.

Second: the memory engine is the feature with the widest gap between promise and maturity. The vendor describes continuous learning from your work; in my testing the context it surfaced was useful mainly for recurring references (project names, prior notes) rather than deep recall of past decisions. It helped, but I would not describe it as a knowledge system yet. Relatedly, the interface is dense — first-run involves workspaces, assistants, policies, memory settings and a browser all at once, and third-party reviewers independently flag the learning curve. Budget an afternoon, not five minutes, for a team rollout.

Third: the model access itself is the make-or-break line item. One seat at $29/month (annual) covering GPT, Claude and Gemini families is genuinely cheaper than three separate subscriptions — that arithmetic is real, and for teams that actually use multiple models it is the reason to buy. But the flip side is what you give up: you are using models through Floutwork's wrapper, without the native apps' ecosystem features (custom GPTs, Projects, artifacts-style canvases, operator features), and no full model list is published — the pricing calculator names Perplexity and Grok but does not enumerate versions. If your team lives inside one vendor's ecosystem, this is a downgrade dressed as savings.

Fourth: product maturity signals are mixed. The pricing history is turbulent — older listings and reviews describe a free forever tier, $12–$20/month personal plans, and an $89 lifetime deal, all of which are gone from the current site; Capterra still shows a $12 plan and the help documentation a 14-day trial against the official 7-day one. The team-product direction is clearly the focus now, and the admin features (SSO/SCIM/SAML on Pro, usage controls, workspace policies) are real and appropriately targeted at 10–1,000-seat rollouts.

Where it stops

• No permanent free tier. The free-forever plan of the solo era is retired; today it is a 7-day trial, then $29/user/month (annual).
• Browser compatibility is the migration blocker. It is a custom browser, and users who depend on specific Chrome extensions — developer tooling, accessibility add-ons — may find gaps; full migration is not practical for everyone.
• No official mobile app. The apps are desktop only as far as I could verify; some directories list mobile versions that I could not confirm on official pages.
• Model transparency is thin. No published model/version list, and you cannot bring your own API keys to verify what runs your prompts.
• Documentation and independent reviews are sparse. A small review base and thin docs make pre-purchase diligence harder than with established players.
• Ecosystem integrations are underspecified. Slack, Jira and similar workflows appear in marketing and assistant examples, but there is no public integration directory I could find.

Verdict

Floutwork is an ambitious consolidation play aimed at the team that pays for four AI subscriptions and loses hours to tab sprawl. Where it works — workspaces, tab grouping, tasks and notes beside the page they came from, one seat for several model families, real admin governance — it earns its keep for a 10-to-1,000-person rollout with mixed model usage. Where it stumbles — dense onboarding, immature memory, unverifiable model versions, retired free tier, thin docs — it stumbles in exactly the places enterprise buyers probe. Balancing depth of core functionality at 40%, value at 20%, usability at 20%, support and ecosystem at 10% and reliability signals at 10%, I rate Floutwork 3.3. For a team standardizing AI access with governance, run the 7-day trial against your real workflows and check the model list with sales before committing; for a solo user with one favorite model, the native apps remain the better deal.

One final note on fit. The buyer I keep coming back to is the department head whose team has quietly accumulated five AI subscriptions with no shared policy, no usage visibility and no onboarding path for new hires. Floutwork solves that specific mess well enough to justify a trial, and the governance surface is the part competitors in this space tend to skip. Whether the workspace itself sticks is a taste question — some people will love one app for everything, others will resent leaving Chrome — so run the trial with both kinds of user before standardizing.

Sources

• https://www.floutwork.com/
• https://floutwork.com/pricing
• https://www.floutwork.com/features
• https://www.floutwork.com/security
• https://floutwork.mintlify.app/introduction
• https://bestsitestobuy.com/ai-tools/floutwork
• https://rightaichoice.com/tools/floutwork
• https://switchtools.io/tool/floutwork
• https://gistdam.com/floutwork
• https://aiportalx.com/tools/floutwork