DeepSeek V3.2

JSON →
deepseek llm
text

An updated version of DeepSeek V3 with improved performance and efficiency for general text and code tasks.

context window 164K tokens
max output 164K tokens
input price $0.28 / 1M tokens
output price $0.4 / 1M tokens
streamingcode-generationreasoningfunction-callingtool-useprompt-caching
releasedMar 2025
full doc /v1/models/deepseek-v3.2