The TP5 models. Built by our lab.

TP5 Mini, TP5 Pro and TP5 Max are the Thinking Platforms AI models. They power FlowOS agents, from quick summaries to the most demanding analyses.

Depending on the setup agreed with you: open-source models hosted in Europe or leading market models.

Request a demo →Discover the lab

TP5 MiniFast

The fastest and most cost-efficient, for everyday tasks.

  • Summaries and meeting minutes
  • Email and customer service replies
  • Sorting and filing documents

TP5 ProVersatile

The everyday model: it adapts its reasoning to each task and each chat mode.

  • Writing deliverables and proposals
  • Document and contract analysis
  • Agent chains and scheduled tasks

TP5 MaxIn-depth

The deepest reasoning, for complex cases and important decisions.

  • Due diligence and financial analysis
  • Strategy and scenarios
  • Long files, multiple sources

Compare the models.

TP5 MiniTP5 ProTP5 Max
Speed
Reasoning depth
Consumption

More dots means a higher level. Consumption follows actual usage: you pay for what you use, no less, no more.

Five generations since 2023.

TP models are developed by the Thinking Platforms research lab. Each generation builds on feedback from the previous one.

2023TP1

Our lab’s first model.

TP2

Second generation.

TP3

Third generation.

TP4

Fourth generation.

TodayTP5

A family of three models: TP5 Mini, TP5 Pro and TP5 Max.

How the TP5 models work.

The right engine for every task

Our lab continuously evaluates AI engines and routes each request to the best fit, based on the TP5 model selected and the nature of the task.

In the FlowOS chat

By default, FlowOS uses the model recommended for the agent. You can also pick TP5 Mini, Pro or Max in the Model menu, and combine it with the Fast, Plan, In-depth or Deliverable modes.

Controlled consumption

Budgets and caps per agent, cost and time-saved tracking on the dashboard: you choose the model with full visibility.

Frequently asked questions.

What is the difference between TP5 Mini, TP5 Pro and TP5 Max?

TP5 Mini favors speed and cost for everyday tasks, TP5 Pro is versatile and adapts to each task, and TP5 Max offers the deepest reasoning for complex cases.

Can I choose the model?

Yes. In the FlowOS chat, the Model menu offers the recommended model by default, the three TP5 models and the other models connected to your workspace.

Who develops the TP models?

The Thinking Platforms research lab, the AI research arm of the Laystone Group. TP models have existed since 2023: TP1, TP2, TP3, TP4, then TP5.

How is usage billed?

Based on consumption: your bill follows your teams' actual usage, with budgets and caps per agent.

Try the TP5 models on your documents.

A 30-minute demo, tailored to your roles.

Request a demo →