Pro
DeepSeek V3.2
Flagship DeepSeek model for coding, reasoning, and agentic work
Cheapest input
$0.27
$/MTok · models.dev
Cheapest output
$0.42
$/MTok
Context
164K
tokens
Offers
1
1 source
Offers
Same model, different listings. First-party, router, and aggregator rows stay separate so you can see the spread instead of a single blended number.
| Source | Provider | Input | Output | Cache read | Context |
|---|---|---|---|---|---|
| models.devcheapestPro/deepseek-ai/DeepSeek-V3.2 | SiliconFlow (China) | $0.27 | $0.42 | $0 | 164K |
Capabilities
- Reasoning
- Tool calling
- Structured output
- In: text
- Out: text
Released: 2025-12-03Family: deepseek
Scores
Sparse coverage on purpose. Missing a score means we have not linked one yet, not that the model is unranked. Full boards on /benchmarks.
- AA intelligence32
- AA coding44.2
- AA agentic18.3
Sources: Arena AI, Artificial Analysis. See /benchmarks.