Skip to content
Tech News
← Back to articles

DeepSeek V4 Peak Valley Pricing Change

read original more articles
Why This Matters

DeepSeek V4's introduction of a peak-valley pricing model highlights a shift towards more dynamic and cost-sensitive API usage, reflecting broader industry trends in optimizing resource allocation and pricing strategies. This change impacts developers and businesses by requiring them to plan API usage around peak hours to manage costs effectively. The update underscores the importance of flexible pricing models in the evolving AI and data processing landscape.

Key Takeaways

DeepSeek V4 Launches in Mid-July with Peak-Valley Pricing

On-chain news: DeepSeek V4 launches in mid-July with a new peak-valley pricing model. API costs double during peak hours (9:00–12:00 and 14:00–18:00 Beijing time). For deepseek-v4-pro, regular pricing is 0.025 RMB (cache hit), 3.00 RMB (cache miss), and 6.00 RMB for output, rising to 0.05 RMB, 6.00 RMB, and 12.00 RMB during peak times. The lightweight model, deepseek-v4-flash, charges 0.02 RMB (cache hit), 1.00 RMB (cache miss), and 2.00 RMB for output, doubling to 0.04 RMB, 2.00 RMB, and 4.00 RMB during peak hours. Users will receive email alerts 24 hours before billing changes. This token launch marks a key update for API users.

ME News reports that, as monitored by Beating on June 29 (UTC+8), DeepSeek officially announced that the official release of DeepSeek V4 is scheduled for mid-July, alongside the introduction of a peak-off-peak pricing mechanism. During peak hours—daily from 9:00 to 12:00 and 14:00 to 18:00 Beijing Time—the API pricing will be doubled compared to regular rates. Under the new pricing structure, for the high-performance model deepseek-v4-pro, the regular rate per million tokens is ¥0.025 for input cache hits, ¥3.00 for input cache misses, and ¥6.00 for output. During peak hours, these rates will increase to ¥0.05, ¥6.00, and ¥12.00, respectively. For the lightweight model deepseek-v4-flash, the regular rate per million tokens is ¥0.02 for input cache hits, ¥1.00 for input cache misses, and ¥2.00 for output. During peak hours, these rates will adjust to ¥0.04, ¥2.00, and ¥4.00, respectively. Users will be notified via email 24 hours prior to any actual pricing changes. (Source: BlockBeats)