Enable Thinking Mode for Advanced Problem Solving and Math
Switch to the reasoner model to access high-compute thinking capabilities.
Toggle your API model parameter to 'deepseek-reasoner' when you need the model to use its internal chain-of-thought for complex problem solving.
The Scenario
You are building a complex software feature or solving a high-level mathematical proof where basic AI response quality isn't sufficient.
Before & after
Manually drafting complex logic or debugging recursive code often takes 45–60 minutes of deep focus and trial-and-error.
Switching to the 'deepseek-reasoner' endpoint takes seconds and activates the model's high-compute 'thinking mode' for deep logic. This takes 1 minute to implement.
The Prompt
// Use this for complex logic or coding tasks in your terminal or API client
model: "deepseek-reasoner",
messages: [
{"role": "user", "content": "Explain the step-by-step logic and provide high-performance code for [INSERT_COMPLEX_PROBLEM_OR_MATH_QUERY]"}
]DeepSeek-V3.2 supports two modes via the API: 'deepseek-chat' for fast, non-thinking responses and 'deepseek-reasoner' for advanced reasoning tasks like coding and math. The latest version uses a hybrid architecture to balance efficiency and depth.
Source
Change Log | DeepSeek API Docs"deepseek-reasoner corresponds to DeepSeek-V3.2's thinking mode"