One account, more than seventy business and AI tools, and a browser for the team beside a Model Context Protocol surface for the agents. The record of the work stays where the work is done.
A team ordinarily assembles its operating system from a chat, a task board, a signing service, a file store, a form builder, several AI subscriptions. Each has its own login, billing, permissions and record.
Raq puts those jobs inside one account and gives compatible agents a common route into them, so a task can begin with a message, use a file, produce a document, request approval and finish on a board — with the whole chain visible to the team.
People and agents work against the same records. Role-based access, audit trails and account switching keep every project inside a single, inspectable perimeter.
Compatible clients — Supercomputer, Codex, Claude Code and others — reach structured operations over the Model Context Protocol, without simulating a browser.
Ask My Agent, Voice Calls and other outward-facing operations take approvals in the browser. The team sees the request, the decision and the outcome side by side.
Every plan includes the entire public catalogue. AI usage and service limits scale with the plan — the tools themselves do not move behind a paywall.
Channels, boards, notes, files and the daily record that keeps a team synchronised.
Every step between a first enquiry and a closed engagement, with a portal the client can actually use.
Drafting, signature, redlining, conversion — the paperwork side of running a business.
Reason across many models. Compare answers. Publish an evaluation. Keep the good prompts.
Media production for the account — with review, batching and brand consistency built in.
Publish it, watch it, understand it — from a site build to Companies House data.
Also includes personal & specialist tools — Compliance Tracker, Spreadsheet AI, Meal Planner, Health Insights.
Open the full indexSupercomputer is Raq's built-in agent. It reads the request in ordinary language, picks the platform tools that fit, works against the same account, files and team as any connected external client, and leaves the result where the team can inspect it.
Raq exposes its tools to compatible agents through the Model Context Protocol. An agent lists boards, creates a card, uploads a file, runs an image batch or reads a project record over a first-class interface — with scoping and approvals attached to the connection.
Because the tools share an account, a task can pass through six of them without leaving a trace behind. The example below shows a client onboarding sequence.
Form Builder captures the prospect and its answers.
Automations create or update the CRM lead.
Task Board opens the onboarding card with owners.
Client Portal opens a branded, magic-link workspace.
E-Sign sends the agreement and captures the audit.
Team Comms carries the internal thread forward.
An agent can inspect the chain end-to-end and prepare anything missing.
The commercial responsibility, runtime and support model differ. The distinction matters — so we keep the two routes named apart.
A customer's coding agent builds a sandboxed application that runs inside the Custom Tools system. Draft visibility for the creator, published visibility for the account, scoped outbound access back into Raq.
A first-class module inside Raq, designed, built, hosted and maintained for the account. First mock-up normally within one working day; a working module within five working days or less after approval, with larger jobs identified during the mock-up stage.
Both routes sit beside the shared tools. Responsibility for the build stays clear either way. A module covers one area of work; a second area is a second module.
Every public tool is included in every paid plan. AI usage and service limits rise with the plan. Prices exclude VAT where applicable. Annual billing currently receives a 5% discount.
A first look at the whole catalogue, with light AI usage and monitoring.
More AI, more monitoring, and the point at which Voice Calls becomes available.
For teams whose work is largely agent-led — production, monitoring and orchestration at scale.
The full envelope for organisations running long agent operations across many accounts.
Custom Modules sit outside the normal plan price at £190 or $259 per module per month. Their AI features draw from the account's normal AI allowance.
Core text and chat requests use Zero Data Retention routes by default. Image, video and voice tools operate under specialist-provider terms — so the two are described apart rather than lumped together.
Vu is an AI and software company based in Solihull. Vu uses Raq for its own work — and hand-builds the managed Custom Modules that customers subscribe to.
Sets the direction of Raq and leads Vu's commercial work with the customers who use it.
Leads the product and engineering work behind the platform, the agent surface and the module practice.
Raq does not train models on customer content.
No card required. One use of each AI tool during the trial. Upgrade whenever.