Kimi K2 Thinking
Kimi K2 Thinking is the reasoning variant of Moonshot AI's Kimi K2 trillion-parameter mixture-of-experts model, tuned to produce extended chain-of-thought before answering. It targets math, coding, and multi-step planning where deliberation improves accuracy, while retaining K2's long-context and bilingual strengths. Use it for hard reasoning tasks, agentic planners, and tool-use loops that benefit from explicit thinking traces.
Pricing
Input
$0.600 / 1M
Output
$2.50 / 1M
Pay only for what you use. No subscriptions, no minimums.
Specs
- Context
- 128K
- Max output
- 98K
- Latency
- Slow
- Category
- Text
- Arabic
- Partial
Quick start
Use any OpenAI-compatible client. Just change base_url and your key.
from openai import OpenAI
client = OpenAI(
api_key="tl-xxxxxxxxxxxxxxxxxxxxxxxx",
base_url="https://api.thalam.ai/v1",
)
response = client.chat.completions.create(
model="moonshotai/kimi-k2-thinking",
messages=[
{"role": "user", "content": "Hello from Thalam!"}
],
)
print(response.choices[0].message.content)Similar models
GPT-5.6 Sol
The top tier of OpenAI's GPT-5.6 generation. Frontier reasoning, coding, and agentic work with a 1M-token context and image input.
GPT-5.6 Terra
The balanced tier of OpenAI's GPT-5.6 generation. Same 1M-token context and image input, tuned for everyday production work.