All models
DeepSeek
DeepSeek V3.2
The all-purpose DeepSeek chat model
Text
Reasoning
Tools
- Type
- Chat
- Context
- 164K tokens
- Russian
- high
- Input price
- 21,52 ₽ / 1M tokens
- Output price
- 32 ₽ / 1M tokens
- Release date
- December 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: "deepseek-chat", messages: [{ role: "user", content: "Hi!" }], stream: true,})for await (const chunk of stream) { process.stdout.write(chunk.choices[0]?.delta?.content ?? "")}About the model
DeepSeek V3.2 is a mature all-purpose model for everyday product tasks: conversations and assistants, text generation and editing, working with code and structured data. The model follows instructions and system prompts predictably, keeps the context of a long conversation accurately and answers confidently in Russian. It supports tool calling (function-calling) and suits agents and integrations where consistent answers and reasonable cost matter more than maximum reasoning depth.
Strengths
- Consistent, predictable quality in conversation
- Code generation and refactoring, working with structured data
- Tool calling (function-calling) for agents and integrations
- Precise adherence to instructions and system prompts
- Inexpensive for the quality of its answers
Where to use it
- Chatbots and first-line support assistants
- Content generation and editing: texts, emails, descriptions
- Help with code: suggestions, explanations, refactoring
- Extracting and structuring data from text
- Automating routine work and agent scenarios with tool calling
- Summarizing, translating and normalizing long documents