Author's Description
DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism designed to improve training and inference efficiency in long-context scenarios while maintaining output quality. Users can control the reasoning behaviour with the `reasoning` `enabled` boolean. [Learn more in our docs](https://openrouter.ai/docs/use-cases/reasoning-tokens#enable-reasoning-with-default-config) The model was trained under conditions aligned with V3.1-Terminus to enable direct comparison. Benchmarking shows performance roughly on par with V3.1 across reasoning, coding, and agentic tool-use tasks, with minor tradeoffs and gains depending on the domain. This release focuses on validating architectural optimizations for extended context lengths rather than advancing raw task accuracy, making it primarily a research-oriented model for exploring efficient transformer designs.
Key Specifications
Supported Parameters
This model supports the following parameters:
Features
This model supports the following features:
Performance Summary
DeepSeek-V3.2-Exp, an experimental model from DeepSeek, demonstrates moderate speed performance, ranking in the 36th percentile across benchmarks. It offers cost-effective solutions, placing in the 64th percentile for price. Notably, the model exhibits exceptional reliability with a 100% success rate, consistently providing usable responses without technical failures. In terms of benchmark performance, DeepSeek-V3.2-Exp shows strong capabilities across various domains. It achieves high accuracy in Coding (93.0%), General Knowledge (99.5%), and Mathematics (94.0%), indicating robust understanding and problem-solving skills. Its Instruction Following (71.7%) and Reasoning (86.0%) scores are also commendable, suggesting effective processing of complex directives and logical tasks. The model excels in Ethics with a perfect 100% accuracy, making it the most accurate and efficient model at its price point for this category. Hallucinations are well-managed with 96.0% accuracy, and Email Classification is strong at 98.0%. While its primary focus is architectural optimization for long contexts, the model maintains performance comparable to V3.1, with minor trade-offs and gains. Its key strength lies in its high accuracy across critical cognitive tasks and its perfect reliability, making it a valuable research-oriented model for exploring efficient transformer designs.
Model Pricing
Current Pricing
| Feature | Price (per 1M tokens) |
|---|---|
| Prompt | $0.21 |
| Completion | $0.32 |
Price History
Available Endpoints
| Provider | Endpoint Name | Context Length | Pricing (Input) | Pricing (Output) |
|---|---|---|---|---|
|
DeepSeek
|
DeepSeek | deepseek/deepseek-v3.2-exp | 131K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
Novita
|
Novita | deepseek/deepseek-v3.2-exp | 163K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
DeepInfra
|
DeepInfra | deepseek/deepseek-v3.2-exp | 163K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
GMICloud
|
GMICloud | deepseek/deepseek-v3.2-exp | 163K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
AtlasCloud
|
AtlasCloud | deepseek/deepseek-v3.2-exp | 163K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
Novita
|
Novita | deepseek/deepseek-v3.2-exp | 163K | $0.27 / 1M tokens | $0.41 / 1M tokens |
|
SiliconFlow
|
SiliconFlow | deepseek/deepseek-v3.2-exp | 163K | $0.27 / 1M tokens | $0.41 / 1M tokens |
|
DeepInfra
|
DeepInfra | deepseek/deepseek-v3.2-exp | 163K | $0.21 / 1M tokens | $0.32 / 1M tokens |
|
Chutes
|
Chutes | deepseek/deepseek-v3.2-exp | 131K | $0.21 / 1M tokens | $0.32 / 1M tokens |
Benchmark Results
| Benchmark | Category | Reasoning | Strategy | Free | Executions | Accuracy | Cost | Duration |
|---|
Other Models by deepseek
|
|
Released | Params | Context |
|
Speed | Ability | Cost |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 Speciale | Dec 01, 2025 | — | 131K |
Text input
Text output
|
★ | ★★★★★ | $$$$ |
| DeepSeek: DeepSeek V3.2 | Dec 01, 2025 | — | 131K |
Text input
Text output
|
— | — | $$$ |
| DeepSeek: DeepSeek V3.1 Terminus | Sep 22, 2025 | ~671B | 131K |
Text input
Text output
|
★★★★ | ★★★★★ | $$$$ |
| DeepSeek: DeepSeek V3.1 Terminus (exacto) | Sep 22, 2025 | ~671B | 131K |
Text input
Text output
|
— | — | $$$ |
| DeepSeek: DeepSeek V3.1 | Aug 21, 2025 | ~671B | 131K |
Text input
Text output
|
★★ | ★★★★ | $$$ |
| DeepSeek: DeepSeek V3.1 Base Unavailable | Aug 20, 2025 | ~671B | 163K |
Text input
Text output
|
★ | ★ | $$ |
| DeepSeek: R1 Distill Qwen 7B Unavailable | May 30, 2025 | 7B | 131K |
Text input
Text output
|
★ | ★ | $$$$ |
| DeepSeek: DeepSeek R1 0528 Qwen3 8B Unavailable | May 29, 2025 | 8B | 131K |
Text input
Text output
|
★★★ | ★★★ | $$ |
| DeepSeek: R1 0528 | May 28, 2025 | ~671B | 128K |
Text input
Text output
|
★★★ | ★★★ | $$$ |
| DeepSeek: DeepSeek Prover V2 | Apr 30, 2025 | ~671B | 131K |
Text input
Text output
|
★★ | ★★★★ | $$$$ |
| DeepSeek: DeepSeek V3 Base Unavailable | Mar 29, 2025 | ~671B | 163K |
Text input
Text output
|
★ | ★ | $$$ |
| DeepSeek: DeepSeek V3 0324 | Mar 24, 2025 | ~685B | 163K |
Text input
Text output
|
★★★★ | ★★★★★ | $$ |
| DeepSeek: R1 Distill Llama 8B Unavailable | Feb 07, 2025 | 8B | 32K |
Text input
Text output
|
★ | ★★ | $$ |
| DeepSeek: R1 Distill Qwen 1.5B Unavailable | Jan 31, 2025 | 5B | 131K |
Text input
Text output
|
★★★ | ★ | $$$ |
| DeepSeek: R1 Distill Qwen 32B | Jan 29, 2025 | 32B | 131K |
Text input
Text output
|
★ | ★★★★ | $$$ |
| DeepSeek: R1 Distill Qwen 14B Unavailable | Jan 29, 2025 | 14B | 32K |
Text input
Text output
|
★ | ★★ | $$$ |
| DeepSeek: R1 Distill Llama 70B | Jan 23, 2025 | 70B | 131K |
Text input
Text output
|
★★★ | ★★★★★ | $$ |
| DeepSeek: R1 | Jan 20, 2025 | ~671B | 128K |
Text input
Text output
|
★★★ | ★★★★ | $$$ |
| DeepSeek: DeepSeek V3 | Dec 26, 2024 | — | 163K |
Text input
Text output
|
★★★ | ★★★★ | $$$ |