DeepSeek V4 Pro vs DeepSeek V4 Flash
Detailed comparison of API pricing, technical specifications, and benchmark scores. Updated 2026-06-29.
Cost at a Glance
DeepSeek V4 Pro is 3.1× more expensive than DeepSeek V4 Flash — switching could save 67.9% on your monthly API bill.
At 5M input + 2M output tokens per month:
Choosing DeepSeek V4 Flash could save you $2.66/mo
Price Comparison (per 1M tokens)
| Metric | DeepSeek V4 Pro | DeepSeek V4 Flash |
|---|---|---|
| Input (per 1M tokens) | $0.43 | $0.14 |
| Output (per 1M tokens) | $0.87 | $0.28 |
| Cache Write (per 1M tokens) | $0.43 | $0.14 |
| Cache Read (per 1M tokens) | $0.00 | $0.00 |
| Batch Input (per 1M tokens) | — | — |
| Batch Output (per 1M tokens) | — | — |
Prices are per 1 million tokens. Actual costs depend on your monthly token volume. Use the calculator →
Specifications
| Spec | DeepSeek V4 Pro | DeepSeek V4 Flash |
|---|---|---|
| Context Window | 1M | 1M |
| Max Output Tokens | 128K | 64K |
| Knowledge Cutoff | 2026-05 | 2026-05 |
| Modalities | Text | Text |
| Prompt Caching | No | No |
| Batch API | No | No |
| Thinking / Reasoning | Yes | No |
Benchmark Scores
| Benchmark | DeepSeek V4 Pro | DeepSeek V4 Flash |
|---|---|---|
| SWE-bench Verified | 80.6 | 79.0 |
| MMLU-Pro | 84.0 | 80.5 |
| HumanEval | 93.8 | 91.2 |
Benchmark scores from official provider reports. Higher is better. Missing scores indicate the provider has not published that benchmark yet.
Use Cases
DeepSeek V4 Pro
DeepSeek V4 Flash
About the Providers
DeepSeek V4 Pro is developed by DeepSeek. DeepSeek V4 Flash is developed by DeepSeek. Both models are from the same provider, representing different tiers in the DeepSeek model lineup. For the most up-to-date pricing, always check the official API documentation from each provider.