TYPHOON Logo
TYPHOON
TYPHOON T1

TYPHOON T1

Reasoning
Llama
3B

First open reasoning model in Southeast Asia with structured thinking paradigm and Thai reasoning traces.

Released
January 23, 2025
Context
N/A
Input
Text
Output
Text
TYPHOON T1
About this Model

[Research Preview] The first open reasoning model in Southeast Asia is here! TYPHOON T1 3B, the debut model in our "TYPHOON T" series, is setting a new benchmark for structured, thoughtful AI reasoning—excelling in math, coding, and other complex tasks.

Key Features
First Open Reasoning Model in Southeast Asia
TYPHOON T1 3B is the first in the TYPHOON T family, showing improved performance across challenging benchmarks compared to the original TYPHOON 2 3B Instruct.
Fast & Low-Compute Requirements
Offers a fast, low-compute requirements model, yet is capable in a variety of tasks by scaling test-time compute, enabling the model to think longer before giving a final answer.
Cross-Domain Reasoning
Able to reason across domains, unlike many open reasoning models limited to mathematics and coding.
Structured Thinking Paradigm
Introduces a new thinking paradigm for reasoning models, structured thinking, where auxiliary tokens help structure the thinking process of the model.
Thai Reasoning Traces
The first reasoning model intentionally equipped with the ability to generate Thai reasoning traces, improving transparency and interpretability.
Release History
Version 2
February 1, 2025
Adding the ability to 🇹🇭 generate Thai reasoning traces, improved Thai performance in general, and enhanced instruction following. This version has the a comparative level of English performance to v2025-01-23.
Version 1
January 23, 2025
Initial release