Happy Horse 1.0 is now on ModelsLab

Try Now
Skip to main content
Available now on ModelsLab · Language Model

Typhoon 2.1 12BThai-English Powerhouse

Try Typhoon 2.1 12BAPI Documentation

Deploy Typhoon 2.1 12B Now

Bilingual Core

Thai-English Mastery

Handles Thai and English queries with Gemma3 12B base for precise bilingual output.

Reasoning Toggle

Switch Thinking Mode

Enable thinking mode via enable_thinking parameter for advanced reasoning tasks.

API Compatible

OpenAI-Style Endpoint

Use Typhoon 2.1 12B API with standard chat completions, 56K tokens context.

Examples

See what Typhoon 2.1 12B can create

Copy any prompt below and try it yourself in the playground.

Recipe Query

“You are a helpful cooking assistant. Provide a step-by-step recipe for Thai green curry in Thai language.”

Code Debug

“Debug this Python function that calculates Fibonacci numbers and optimize it for speed: def fib(n): if n <= 1: return n else: return fib(n-1) + fib(n-2)”

Math Proof

“Prove that the sum of angles in a triangle is 180 degrees. Explain step-by-step with diagrams in text.”

Story Outline

“Outline a sci-fi story about AI exploring ancient Thai ruins, 500 words, in English.”

For Developers

A few lines of code.
Chat Completions. One Call.

ModelsLab handles the infrastructure: fast inference, auto-scaling, and a developer-friendly API. No GPU management needed.

  • Serverless: scales to zero, scales to millions
  • Pay per token, no minimums
  • Python and JavaScript SDKs, plus REST API
API Documentation 
import requests


response = requests.post(
    "https://modelslab.com/api/v7/llm/chat/completions",
    json={
  "key": "YOUR_API_KEY",
  "prompt": "",
  "model_id": ""
}
)
print(response.json())

FAQ

Common questions about Typhoon 2.1 12B

Read the docs

Typhoon 2.1 12B is a 12B parameter instruct LLM based on Gemma3, optimized for Thai-English bilingual tasks. It supports reasoning with a thinking mode toggle. Outperforms larger predecessors in Thai benchmarks.

Use OpenAI-compatible endpoints via Together.ai or OpenTyphoon.ai. Pass model ID typhoon-v2.1-12b-instruct in requests. Authenticate with API key from playground.

Supports 56K tokens for input and output combined. Max tokens default 150, up to 8192 via max_tokens parameter.

Yes, Typhoon 2.1 12B outperforms Typhoon 2 70B in Thai MT-Bench at 1/6 size. Ideal for efficient deployments.

Load via Hugging Face Transformers, vLLM, or Ollama with scb10x/typhoon2.1-gemma3-12b. Use bfloat16 for inference.

5 requests per second, 200 per minute on OpenTyphoon. Adjust temperature 0-2 for output control.

Ready to create?

Start generating with Typhoon 2.1 12B on ModelsLab.

Try Typhoon 2.1 12BAPI Documentation