DeepSeek V3.2 Speciale
Intelligence
29.4
#169/523
Coding
37.9
#67/429
Math
96.7
#5/265
Speed
—
Pricing
—
DeepSeek-V3.2-Speciale is a specialized variant of DeepSeek-V3.2, optimized for enhanced performance on specific tasks.
DeepSeek V3.2 Speciale is DeepSeek’s latest model designed for high-performance tasks. It excels in mathematical computations with a Math Index of 96.7, making it suitable for applications requiring advanced calculations. This model targets professional users in fields such as data analysis and scientific research.
When to Use DeepSeek V3.2 Speciale
✓ Best For
- Complex mathematical modeling
- Data analysis in scientific research
- High-performance coding tasks
✗ Not Ideal For
- General-purpose applications
- Basic coding tasks
How DeepSeek V3.2 Speciale Compares
Intelligence Index · Higher is better
GoogleAlibabaDeepSeekBaiduxAI
Benchmark Profile
Coding Index
DeepSeekOpenAIGoogleNVIDIA
Math Index
GoogleOpenAIDeepSeekXiaomi
Intelligence · Coding · Math
Intelligence
Coding
Math
All Benchmark Scores (12)
| Benchmark | Score | |
|---|---|---|
| Intelligence Index | 29.4 | |
| Coding Index | 37.9 | |
| Math Index | 96.7 | |
| MMLU-Pro | 863% | |
| GPQA | 871% | |
| LiveCodeBench | 896% | |
| HLE | 261% | |
| SciCode | 44% | |
| IFBench | 63.9% | |
| LCR | 59.3% | |
| TerminalBench Hard | 34.8% | |
| AIME 2025 | 96.7% |
Data: Artificial Analysis · Updated: April 9, 2026
Frequently Asked Questions (15)
When was DeepSeek V3.2 Speciale released?
DeepSeek V3.2 Speciale was released on December 1, 2025.
Who created DeepSeek V3.2 Speciale?
DeepSeek V3.2 Speciale was created by DeepSeek.
How intelligent is DeepSeek V3.2 Speciale?
DeepSeek V3.2 Speciale scores 29 on the Artificial Analysis Intelligence Index, placing it above average among other open weight models of similar size (median: 27).
How verbose is DeepSeek V3.2 Speciale?
When evaluated on the Intelligence Index, DeepSeek V3.2 Speciale generated 110M output tokens, which is at the higher end compared to other open weight models of similar size (median: 39M).
Is DeepSeek V3.2 Speciale a reasoning model?
Yes, DeepSeek V3.2 Speciale is a reasoning model. It uses extended thinking or chain-of-thought reasoning to work through complex problems before providing an answer.
What input modalities does DeepSeek V3.2 Speciale support?
DeepSeek V3.2 Speciale supports text input.
What output modalities does DeepSeek V3.2 Speciale support?
DeepSeek V3.2 Speciale supports text output.
Can DeepSeek V3.2 Speciale process images?
No, DeepSeek V3.2 Speciale does not support image input. It can only process text.
Is DeepSeek V3.2 Speciale multimodal?
No, DeepSeek V3.2 Speciale is not multimodal. It only supports text input.
What is the context window of DeepSeek V3.2 Speciale?
DeepSeek V3.2 Speciale has a context window of 130k tokens. This determines how much text and conversation history the model can process in a single request.
Is DeepSeek V3.2 Speciale open source?
Yes, DeepSeek V3.2 Speciale is open weights. The model weights are publicly available and can be downloaded for self-hosting.
How many parameters does DeepSeek V3.2 Speciale have?
DeepSeek V3.2 Speciale has 685 billion parameters (37 billion active).
What are the active parameters of DeepSeek V3.2 Speciale?
DeepSeek V3.2 Speciale is a Mixture of Experts (MoE) model with 685 billion total parameters, but only 37 billion active parameters are used during inference.
What is the license for DeepSeek V3.2 Speciale?
DeepSeek V3.2 Speciale is released under the MIT license. This license allows commercial use.
How does DeepSeek V3.2 Speciale perform on benchmarks?
DeepSeek V3.2 Speciale achieves a score of 29 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.
