DeepSeek-V3.2 Speciale API

DeepSeek V3.2 Speciale is an advanced reasoning model with a thinking-only mode.
Output

How to use DeepSeek-V3.2 Speciale API

Install any OpenAI-compatible SDK, point it at api.aimlapi.com/v1, and set the model to .

OpenAI-compatible — swap the base URL and it works with your existing SDK.

DeepSeek-V3.2 Speciale API Pricing

TypePrice
Input
Output

DeepSeek-V3.2 Speciale Benchmarks

BenchmarkScoreWhat it measuresSourceRetrieved
Intelligence
14.5
Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial AnalysisSourceSeptember 12, 2026
Math
96.7
Composite score across standardised mathematics evaluations, measured independently by Artificial AnalysisSourceSeptember 12, 2026

DeepSeek-V3.2 Speciale vs other models

ModelInputOutputContextBest for
$6.5 / 1M tokens
$39 / 1M tokens
1.05M tokens
Reasoning + agents
$2.6 / 1M tokens
$13 / 1M tokens
1M tokens
Balanced coding + agents
$3.9 / 1M tokens
$19.5 / 1M tokens
1M tokens
Long-context, multimodal & agentic workflows
$0.65 / 1M tokens
$3.9 / 1M tokens
1.05M tokens
Reasoning + agents

Frequently asked questions

Yes, DeepSeek-V3.2 Speciale can stream responses as they are generated.

Yes, DeepSeek-V3.2 Speciale supports both function calling and structured outputs.

DeepSeek-V3.2 Speciale was built by DeepSeek.

It supports function calling, streaming, and structured outputs, which are features commonly needed to build agentic workflows.

It is accessed via the v1 endpoint.

Start building with DeepSeek-V3.2 Speciale

Get API Key
1000+ models, one API.