Model specification · DeepSeek

DeepSeek V4 Flash 0731

deepseek/deepseek-v4-flash-0731

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is suited for coding, reasoning, and agent workflows.

SpecificationsReported values
Context window
1.0M
Maximum output
66K
Architecture
text->text
Tokenizer
DeepSeek
Knowledge cutoff
Moderated
No
Input modalities
text
Output modalities
text
Full pricingUSD where reported
PriceStandard
Input$0.090 / M
Output$0.180 / M
Benchmark indexesReported indexes
Intelligence
49.9
Coding
69.1
Agentic
45.7
Design Arena Elo
Example workload costsStandard pricing only
WorkloadInput / outputCache shareEstimated cost
Quick check10K / 1K × 10%$0.00108
Long analysis100K / 4K × 150%$0.00612
Workday20K / 2K × 1,00030%$1.73
More from DeepSeekAll DeepSeek models
ModelContextInput / MOutput / MIntel.
DeepSeek V4 Pro1.0M$0.435$0.87044.3
DeepSeek V3.2164K$0.269$0.40032
DeepSeek V3.1 Terminus164K$0.270$1.0030.4
R1164K$0.700$2.5018.5
DeepSeek V3 0324164K$0.270$1.1215.4