DeepSeek V3.2 by Venice AI — Specs, Pricing & Capabilities | whichllm

Model ID deepseek-v3.2

Provider Venice AI

Family deepseek

Status -

Knowledge Cutoff 2025-10

Release Date 2025-12-04

Input Modalities text

Output Modalities text

Context Window 160000

Input Limit -

Output Limit 32768

Tool Calling Yes

Reasoning Yes

Structured Output Yes

Temperature Control Yes

Open Weights Yes

Input Cost / 1M tokens $0.33

Output Cost / 1M tokens $0.48

Reasoning Cost / 1M tokens -

Cache Read Cost / 1M tokens $0.16

Cache Write Cost / 1M tokens -