AI Flora Competition

I had three flagship AI models compete to find out which one was truly the best at coding.
The three contestants were GLM-5.2, 5.6-Sol, and Sonnet 5.
Of the three, one prevailed.

5.6-Sol

2nd Place

OpenAI's flagship model for advanced reasoning, coding, research, and high-performance agent workflows.

---

See page See result

GLM-5.2

1st Place

Open-weight model built for coding, reasoning, and long-horizon agent tasks with a 1M-token context.

---

See page See result

Sonnet 5

3rd Place

Fast, capable model for coding, reasoning, and everyday AI tasks.

---

See page See result