Pricing
Same rates as the standard page, but the model columns are presented in a different order — the parser must align by the MODEL header row, not by fixed column positions.
| MODEL | deepseek-v4-pro | deepseek-v4-flash-vision-exp | deepseek-v4-flash | ||
| BASE URL (OpenAI Format) | https://api.deepseek.com | ||||
| PRICING(1)(2) | 1M INPUT TOKENS (CACHE HIT) | OFF-PEAK | $0.022 | $0.007 | $0.007 |
| PEAK | $0.044 | $0.014 | $0.014 | ||
| 1M INPUT TOKENS (CACHE MISS) | OFF-PEAK | $0.66 | $0.22 | $0.22 | |
| PEAK | $1.32 | $0.44 | $0.44 | ||
| 1M OUTPUT TOKENS | OFF-PEAK | $1.98 | $0.66 | $0.66 | |
| PEAK | $3.96 | $1.32 | $1.32 | ||
| Concurrency Limit(3) | 500 | 2500 | 2500 | ||
(1) Off-peak rates are half of the peak rates. Peak hours are 01:00 - 04:00 and 06:00 - 10:00 UTC (all other hours are off-peak).