DeepSeek-V4-Flash-0731 vs DeepSeek-V4-Pro-0813: Specs & Benchmark Comparison
DeepSeek-V4-Flash-0731 and DeepSeek-V4-Pro-0813 both come from DeepSeek, DeepSeek-V4-Flash-0731 was released in July 2026, and DeepSeek-V4-Pro-0813 followed a month later in August 2026. DeepSeek-V4-Pro-0813 is the larger model at roughly 1.6 trillion parameters, against 304 billion for DeepSeek-V4-Flash-0731.
The two models share 9 published benchmarks. DeepSeek-V4-Pro-0813 leads on 9 of them. The widest gaps are on DeepSWE, where DeepSeek-V4-Pro-0813 scores 62.7% against 54.4%; DSBench-Hard, where DeepSeek-V4-Pro-0813 scores 67.2% against 59.6%. Averaged across everything we track, DeepSeek-V4-Flash-0731 sits at 57.5% and DeepSeek-V4-Pro-0813 at 60.6%.
DeepSeek-V4-Flash-0731 is the cheaper API at $0.14 per million input tokens and $0.28 per million output tokens, roughly 9 times cheaper than DeepSeek-V4-Pro-0813 at $1.32 and $3.96. Both accept a context window of about 1M tokens.
| Characteristic | DeepSeek-V4-Flash-0731 | DeepSeek-V4-Pro-0813 |
|---|---|---|
| Company | DeepSeek | DeepSeek |
| Release Date | July 31, 2026 | August 13, 2026 |
| Parameters | 304B | 1.6T |
| Multimodal | No | No |
| Context (input) | 1.0M | 1.0M |
| Context (output) | 393K | 393K |
| Input Price / 1M | $0.14 | $1.32 |
| Output Price / 1M | $0.28 | $3.96 |
| Average Score | 57.5% | 60.6% |
| Benchmarks | ||
| DeepSWE | 54.4% | 62.7% |
| DSBench-Hard | 59.6% | 67.2% |
| NL2Repo | 54.2% | 61.5% |
| AutomationBench | 25.1% | 31.8% |
| CyberGym | 77.0% | 83.0% |
| Terminal-Bench 2.1 | 83.0% | 88.0% |
| Toolathlon | 70.0% | 74.0% |
| DSBench-FullStack | 69.0% | 71.0% |
| Agents' Last Exam | 25.2% | 25.7% |
Visual Benchmark Comparison
Verdict
DeepSeek-V4-Flash-0731 leads in 1 out of 4 comparison categories.
Both models show comparable average scores: DeepSeek-V4-Flash-0731 — 0.6, DeepSeek-V4-Pro-0813 — 0.6.
DeepSeek-V4-Flash-0731 is 12.6x cheaper: input $0.14/1M vs $1.32/1M tokens.
Same context size: 1M tokens.
Both models were released around the same time: 7/31/2026 and 8/13/2026.
More About These Models
Related Comparisons
Frequently Asked Questions
Which is better for coding — DeepSeek-V4-Flash-0731 or DeepSeek-V4-Pro-0813?
Which model is cheaper — DeepSeek-V4-Flash-0731 or DeepSeek-V4-Pro-0813?
Which has a larger context window — DeepSeek-V4-Flash-0731 or DeepSeek-V4-Pro-0813?
The DeepSeek-V4-Flash-0731 and DeepSeek-V4-Pro-0813 comparison is updated for 2026. Data includes benchmark results, API pricing, context window size and other specifications. For more detailed information, visit the DeepSeek-V4-Flash-0731 or DeepSeek-V4-Pro-0813 page. See also the complete list of AI model comparisons.