Skip to content

Marco-o1

An open reasoning model with CoT and MCTS

Visit official site

What Marco-o1 is

Marco-o1 is described as an open large reasoning model that targets complex real world problems with several reasoning strategies.

What you can do with it

  • Fine tune with CoT data and synthetic data
  • Expand the solution space with MCTS
  • Use reflection to improve search efficiency

Application scenarios

  • Find a solution to a maths problem
  • Translate slang and spoken expressions
  • Work out an answer to a coding problem

Pros & cons

✓ What we like

  • Open ended problems are addressed alongside standard answer subjects
  • Translation tasks are listed as an application

! What to watch out for

  • Use starts by cloning the GitHub repository
  • Inference loads the Marco-o1-CoT model

FAQ

Who developed it?

The material names the MarcoPolo team of Alibaba International Digital Commerce.

What search method is used?

The features describe Monte Carlo Tree Search guided by output confidence.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review