DeepSeek V3.2
deepseek-v3.2
Intelligence (AA)
41.7
source: Artificial Analysis
Speed
31 t/s
median output throughput
Coding
36.7
Artificial Analysis coding index
Providers
3
with listed pricing
Providers for DeepSeek V3.2
Model facts
- Model ID
- deepseek-v3.2
- Tier
- Efficient
- Context window
- 160K
- Privacy
- Venice — no data retention
Cite: DeepSeek V3.2
Citation
DeepSeek V3.2 — Inference Index, 2026-04-16. Retrieved from https://inferenceindex.ai/models/deepseek-v3-2API
curl https://inferenceindex.ai/api/prices