▶ ai·
Anthropic API
To meet your needs for using the Anthropic API ecosystem, our API now supports the Anthropic API format with a base_url of https://api.deepseek.com/anthropic. With simple configuration, you can integrate DeepSeek's…
items8 latest
▶ ai·
FIM Completion (Beta)
In FIM (Fill In the Middle) completion, users can provide a prefix and suffix (optional), and the model completes the content in the middle. FIM is commonly used for content continuation, code completion, and other…
▶ ai·
Multi-turn dialogue
This guide will introduce how to use the DeepSeek /chat/completions API for multi-turn conversations. The DeepSeek /chat/completions API is a "stateless" API, meaning the server does not record the context of user…
▶ ai·
Conversation Prefix Continuation (Beta)
Conversation prefix continuation uses the Chat Completion API, where users provide messages beginning with assistant to have the model complete the rest of the message. Notes When using conversation prefix…
▶ ai·
Thinking Mode
DeepSeek models support thinking mode: before outputting the final answer, the model will first output a chain of thought content to improve the accuracy of the final answer. Thinking mode toggle and thinking intensity…
▶ ai·
JSON Output
In many scenarios, users need models to output strictly in JSON format to achieve structured output, making it easier for subsequent logic to parse. DeepSeek provides a JSON Output feature to ensure the model outputs a…
▶ ai·
Tool Calls
Tool Calls allow models to invoke external tools to enhance their capabilities. Non-thinking Mode Sample Code Here is a complete Python code example demonstrating the use of Tool Calls, showing how to get weather…
▶ ai·
Context Hard Drive Cache
DeepSeek API context disk caching is enabled by default for all users and requires no code modifications to use. Each user request triggers the construction of disk cache. If subsequent requests have overlapping…