Bordet

Seat Claude, GPT and Gemini in one room with shared memory

Intermediate
Screenshot of Bordet, Seat Claude, GPT and Gemini in one room with shared memory

What is Bordet?

Bordet is a shared AI boardroom where Claude, GPT, Gemini and other models work one project together under assigned roles, with persistent shared memory and an auditable decisions log. Teammates join the same room, and coding agents can take work from it.

Bordet is a multi-model workspace built around a single idea: one chatbot gives you one opinion, so put several models and your teammates at the same table instead. A boardroom is a shared room where each AI member takes a role (a Strategist on Claude, a Maker on GPT, a Critic on Gemini, for example) and works on the same project. The tool is model-agnostic: Claude, GPT, Gemini, Grok, Llama, DeepSeek and Mistral can all sit in the same room, and the human stays the decision-maker while roles change as the work demands. In practice you assemble a board by picking models and assigning roles, or by describing your project and letting Bordet propose the right board from its templates. You can put a question to one member, several, or @everyone. Each answers independently, then members respond to each other's answers, which produces the round-table effect rather than a single threaded reply. A persistent shared memory carries project state, constraints and history across sessions, so context does not have to be re-pasted between tabs. Every decision lands in an append-only decisions log that the whole room can read, so settled questions stay settled. It suits small product, design and analytics teams that already juggle multiple AI subscriptions, and anyone who wants a second and third model to check the first. Teammates join through an invite link and share the same memory and log in real time. Two further routes extend the room: a Claude or ChatGPT subscription can take a seat through MCP, and coding agents such as Claude Code, Codex or Cursor can pick up work from the room, execute it in a repository and record what they did. GitHub and folder sync are included. Among alternatives, Bordet sits between single-provider chat assistants and bare model APIs. It does not build a bespoke agent pipeline; it organises a standing conversation with memory and an audit trail, and it bills either through prepaid credits at a markup over raw model cost or through your own provider keys, with BYOK free during beta.

How do you use Bordet?

  1. 1Open a boardroom
    Start from the homepage and open a boardroom, or sign up free first so your rooms and memory are kept. No card and no API keys are needed for the first session.
    Bordet — Open a boardroom
  2. 2Assemble your AI team
    Choose which models sit at the table and give each one a role, such as a Strategist on Claude or a Critic on Gemini. Mixing providers inside one room is supported.
    Bordet — Assemble your AI team
  3. 3Or start from a project template
    Describe your project and Bordet proposes a board drawn from its templates, which saves assigning roles by hand for common planning work.
    Bordet — Or start from a project template
  4. 4Address the room
    Ask a single member, several, or @everyone. Members answer on their own and then reply to each other, so expect a round-table exchange rather than one reply.
  5. 5Invite your teammates
    Share an invite link so people join the same boardroom alongside the AI members. Everyone reads and writes to the same shared memory and decisions log.
  6. 6Review the decisions log
    Check the append-only record of what was decided and why before the next session, and use it to keep the project state consistent as the work moves on.

Pros and cons

Pros

  • Several frontier models share one room and one memory, so answers get cross-checked instead of taken on trustAI
  • Persistent memory removes the usual re-pasting of project context between separate chatbot tabsAI
  • The append-only decisions log records what was decided and why, which keeps a team from relitigating settled callsAI
  • Prepaid credits cap spending by design, and BYOK rooms pay providers directly at cost with no markupAI
  • Humans and AI members share the same thread, and coding agents can execute work in your own repositoryAI

Cons

  • Credit usage is billed at a markup over raw model cost, so heavy managed use will cost more than plain provider subscriptionsAI
  • BYOK, the cheaper route, is free only while the product is in beta, which makes long-term pricing uncertainAI
  • Model output still depends entirely on the underlying providers, and Bordet gives no offline or local model optionAI
  • Rounding up a room and assigning roles adds setup work that a single chat session does not requireAI
  • No native mobile apps are described, so the boardroom is effectively a browser-bound toolAI

How much does Bordet cost?

Free trial

A free tier gives the full boardroom experience with no card and no API keys, and signing up free keeps your rooms and memory.

Learn more

Pricing

Free tier with no card or keys. Managed use runs on prepaid credits from $5 (packs of $5, $20 or $50) spent at a transparent markup over raw model cost, with the balance acting as a hard cap. BYOK lets you pay OpenAI, Anthropic or Google directly at cost, free while in beta. No subscription or seats.

Learn more

Integrations

Model providers include OpenAI, Anthropic, Google and xAI. Your own Claude or ChatGPT can join via MCP, coding agents such as Claude Code, Codex and Cursor can take work from the room, and GitHub and folder sync are included.

Features

Shared persistent memory across sessions, an auditable append-only decisions log, human and AI members in one real-time thread, per-member roles with named addressing and turn-taking, model-agnostic rooms spanning Claude, GPT, Gemini, Grok, Llama, DeepSeek and Mistral, plus MCP seats for your own Claude or ChatGPT subscription and GitHub and folder sync for coding agents.

Learn more

Frequently asked questions about Bordet

  • How much does Bordet cost?
    Free tier with no card or keys. Managed use runs on prepaid credits from $5 (packs of $5, $20 or $50) spent at a transparent markup over raw model cost, with the balance acting as a hard cap. BYOK lets you pay OpenAI, Anthropic or Google directly at cost, free while in beta. No subscription or seats.
  • Does Bordet offer a free trial?
    A free tier gives the full boardroom experience with no card and no API keys, and signing up free keeps your rooms and memory.
  • How do you use Bordet?
    The walkthrough on this page covers 6 steps: 1. Open a boardroom 2. Assemble your AI team 3. Or start from a project template 4. Address the room 5. Invite your teammates 6. Review the decisions log.
  • What platforms does Bordet support?
    Bordet is available on Web App.
  • What does Bordet integrate with?
    Model providers include OpenAI, Anthropic, Google and xAI. Your own Claude or ChatGPT can join via MCP, coding agents such as Claude Code, Codex and Cursor can take work from the room, and GitHub and folder sync are included.
  • What are the limitations of Bordet?
    Credit usage is billed at a markup over raw model cost, so heavy managed use will cost more than plain provider subscriptions. BYOK, the cheaper route, is free only while the product is in beta, which makes long-term pricing uncertain. Model output still depends entirely on the underlying providers, and Bordet gives no offline or local model option.

Status

StatusActive
Views0
Outbound clicks0
Added10/22/2026

Platforms

Web App

Pricing

FreemiumFree TrialPaid