DeepSeek V3.2 (Reasoning)
DeepSeek V3.2 (Reasoning) is a text-based AI model by DeepSeek. It features 32 tok/s output speed, $0.28/1M input tokens pricing. It scores 41.7 on the intelligence index.
Strengths
- +Solid intelligence score (41.7)
- +Very affordable pricing ($0.28/1M tokens)
Weaknesses
- -Slower than average output (32 tok/s)
Benchmarks
Intelligence41.7Coding36.7Math92.0
MMLU-Pro
0.9
GPQA
0.8
LiveCodeBench
0.9
Intelligence
41.7
Coding
36.7
Math
92.0
Output Speed
32 tok/s
Latency (TTFT)
1.47s
Blended Price
$0.32/M
Pricing per 1M Tokens
| Input (Prompt) | $0.28 |
| Output (Completion) | $0.42 |
| Cache Read | Free |
| Cache Write | Free |
Specifications
| Context Length | N/A |
| Max Output Tokens | N/A |
| Input Modalities | Text |
| Output Modalities | Text |
| Tokenizer | N/A |
Compare DeepSeek V3.2 (Reasoning) with:
Alternatives to DeepSeek V3.2 (Reasoning)
More from DeepSeek
Frequently Asked Questions
How much does DeepSeek V3.2 (Reasoning) cost?
DeepSeek V3.2 (Reasoning) costs $0.28/1M input tokens and $0.42/1M output tokens.
Is DeepSeek V3.2 (Reasoning) good for coding?
DeepSeek V3.2 (Reasoning) has a moderate coding score of 36.7. It can handle basic programming tasks but may not be the best for complex coding.
How fast is DeepSeek V3.2 (Reasoning)?
DeepSeek V3.2 (Reasoning) generates output at 32 tok/s. Time to first token is 1.47s.
Is there a free version of DeepSeek V3.2 (Reasoning)?
No, DeepSeek V3.2 (Reasoning) is a paid model. Check the free models page for zero-cost alternatives.
What are alternatives to DeepSeek V3.2 (Reasoning)?
Last updated: