All models
Moonshot AI
Kimi K2 Thinking
Kimi K2 with a thinking mode
Text
Reasoning
Tools
- Type
- Chat
- Context
- 262K tokens
- Russian
- medium
- Input price
- 48 ₽ / 1M tokens
- Output price
- 200 ₽ / 1M tokens
- Release date
- November 2025
chat/completions
import OpenAI from "openai"const client = new OpenAI({ baseURL: "https://api.yenisei.ru/v1", apiKey: process.env.YENISEI_API_KEY,})const stream = await client.chat.completions.create({ model: "kimi-k2-thinking", messages: [{ role: "user", content: "Hi!" }], stream: true,})for await (const chunk of stream) { process.stdout.write(chunk.choices[0]?.delta?.content ?? "")}About the model
Kimi K2 Thinking is a variant of Kimi K2 with an explicit thinking mode for complex multi-step tasks. The model first works through the task step by step and then answers, which is useful in analysis and agent scenarios. It holds a very long context and suits large documents and correspondence. It calls tools well and carries out multi-step actions, making it a good fit for autonomous agents. It is multilingual and accurate in Russian.
Strengths
- Explicit step-by-step reasoning before answering
- Very long context
- Strong agent scenarios and tool calling
- Planning and coordinating multi-step actions
- Works in many languages, including Russian
- Consistent quality on complex tasks
Where to use it
- Planning and autonomous agents
- Analyzing large documents and correspondence
- Complex analytical and research tasks
- Multi-step dialogues with tools
- Working through long technical materials
- Decision-making with justification