Google LLC (Vertex AI)
deepseek-v3.2
DeepSeek-V3.2 is a model that harmonizes high computational efficiency with superior reasoning and agent performance. DeepSeek's approach is built upon three key technical breakthroughs: DeepSeek Sparse Attention (DSA), scalable reinforcement learning framework, and large scale agentic task synthesis pipeline.
👁Vision🧠Reasoning🔧Tool calling⚡Caching🎨Image gen
Pricing per 1M tokens
Input
$0.56
Output
$1.68
Cache write
$1.68
Cache read
$0.06
Specifications
Context window164K tokens
Max output66K tokens
API typechat
AddedMar 19, 2026
Model IDvertex/deepseek-v3.2
Privacy & data
Data retentionNo
Used for trainingNo
Provider location🇺🇸 US / 🇪🇺 EU
Privacy policyVertex AI Data Governance →