# Distributed Cognition's 0.5M-parameter VINCO beats frontier LLMs at Catan

> Original title: Distributed Cognition’s VINCO, a 0.5M-parameter model beat Claude Opus 5.5, Claude Fable 5.1, Astra and Kimi K3 in 36 of 40 in the the bo...

- Source: Rohan Paul
- Published: 2026-11-05T21:27:19.000Z
- Bad Influence: https://badinfluence.news/items/xwp0ejdwprk4u8xwc1qf51g2e
- Original: https://x.com/rohanpaul_ai/status/2118455143696064635

## Summary

Distributed Cognition says its 0.5M-parameter model VINCO beat Claude Opus 5.5, Claude Fable 5.1, Astra and Kimi K3 in 36 of 40 games of 1v1 Catan. Catan tests planning without seeing the opponent's cards or knowing future dice rolls, which coding benchmarks do not measure. The company says VINCO is about 10 million times smaller than the frontier LLMs it beat.

## Post

Distributed Cognition’s VINCO, a 0.5M-parameter model beat Claude Opus 5.5, Claude Fable 5.1, Astra and Kimi K3 in 36 of 40 in the the board game Catan.

Catan tests a skill coding benchmarks don’t: picking good moves without seeing the opponent’s cards or knowing the next dice roll.

The company’s broader view is that superintelligence will come from many models like this collaborating, rather than one model covering everything. By its numbers, VINCO is about 10 million times smaller than the frontier LLMs it beat in testing.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWZCLIXbUAAY-tk.jpg&mode=thumb&exp=1794182400&sig=612c8bd6af7ec6b4)

## Quoted post @distcog

> We're excited to unveil VINCO, our frontier AI model trained to play and win 1v1 Catan.
> 
> We've secretly deployed it on the http://Colonist.io global leaderboard, where it has reached both the #1 and #2 positions, consistently beating every human world champion.

https://x.com/distcog/status/2118402967808487546