DeepSeek 将涨价归因于需求激增导致的 GPU 算力瓶颈,并成为首家引入结构性峰谷计费(类似电价分时计费)的头部 AI API 提供商 [1][22]。
What new API pricing did DeepSeek announce for its V4 models, when does it take effect, and what are the key details about the models' conteDeepSeek's new peak/off-peak API pricing for V4 models, announced August 13, 2026
AI 提示
Create a landscape editorial hero image for this Studio Global article: What new API pricing did DeepSeek announce for its V4 models, when does it take effect, and what are the key details about the models' conte. Article summary: On August 13, 2026, DeepSeek announced major API price increases for its V4-Pro and V4-Flash models, with new peak/off-peak billing tiers taking effect at **16:00 UTC on August 16, 2026** [1][4]. The increases range from. Topic tags: general, academic, documentation, general web, news. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks,
DeepSeek 将涨价归因于需求激增导致的 GPU 算力瓶颈,并成为首家引入结构性峰谷计费(类似电价分时计费)的头部 AI API 提供商 [1][22]。
What new API pricing did DeepSeek announce for its V4 models, when does it take effect, and what are the key details about the models' conteDeepSeek's new peak/off-peak API pricing for V4 models, announced August 13, 2026
AI 提示
Create a landscape editorial hero image for this Studio Global article: What new API pricing did DeepSeek announce for its V4 models, when does it take effect, and what are the key details about the models' conte. Article summary: On August 13, 2026, DeepSeek announced major API price increases for its V4-Pro and V4-Flash models, with new peak/off-peak billing tiers taking effect at **16:00 UTC on August 16, 2026** [1][4]. The increases range from. Topic tags: general, academic, documentation, general web, news. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks,
DeepSeek 是首家引入结构性分时段动态调价(surge pricing)的头部 AI API 提供商,其模式借鉴了电力领域的峰谷计费 。根据新规,非高峰时段的价格是高峰时段的一半,这意味着高峰期的调用成本是低谷期的两倍 。据报道的 UTC 高峰时段分别为 01:00–04:00 和 06:00–10:00 。
评估自托管方案:两款 V4 模型均为 MIT 许可,可在 Hugging Face 下载。对于高负载工作场景,尽管前期有基础设施成本,但自托管可能带来更优的经济效益 。
重新对比 GPT-5.6 Luna:现在 Luna 的输入/输出价格仅为 $0.20/$1.20。如果您的任务对输出要求高,且不需要 DeepSeek 的 100 万 Token 上下文或特定架构优势,那么 Luna 可能是更便宜的选择。
关键结论
DeepSeek 此次涨价标志着一个时代的终结:它不再是那个价格碾压所有对手的 API 提供商。在某些维度上,其价格已与 OpenAI 最具性价比的模型看齐,但相比 Anthropic 的高端产品仍有巨大优势。对于开发者而言,峰谷计费、深度的缓存折扣以及 MIT 开源权重这三大因素意味着:最终的 API 费用将更多地取决于工作负载架构的巧妙设计,而不仅仅是模型的原始选择。
pandaily.com
DeepSeek V4 Official Version Due Mid-July with Peak-Hour API Pricing Doubling
DeepSeek 是首家引入结构性分时段动态调价(surge pricing)的头部 AI API 提供商,其模式借鉴了电力领域的峰谷计费 。根据新规,非高峰时段的价格是高峰时段的一半,这意味着高峰期的调用成本是低谷期的两倍 。据报道的 UTC 高峰时段分别为 01:00–04:00 和 06:00–10:00 。
评估自托管方案:两款 V4 模型均为 MIT 许可,可在 Hugging Face 下载。对于高负载工作场景,尽管前期有基础设施成本,但自托管可能带来更优的经济效益 。
重新对比 GPT-5.6 Luna:现在 Luna 的输入/输出价格仅为 $0.20/$1.20。如果您的任务对输出要求高,且不需要 DeepSeek 的 100 万 Token 上下文或特定架构优势,那么 Luna 可能是更便宜的选择。
关键结论
DeepSeek 此次涨价标志着一个时代的终结:它不再是那个价格碾压所有对手的 API 提供商。在某些维度上,其价格已与 OpenAI 最具性价比的模型看齐,但相比 Anthropic 的高端产品仍有巨大优势。对于开发者而言,峰谷计费、深度的缓存折扣以及 MIT 开源权重这三大因素意味着:最终的 API 费用将更多地取决于工作负载架构的巧妙设计,而不仅仅是模型的原始选择。
pandaily.com
DeepSeek V4 Official Version Due Mid-July with Peak-Hour API Pricing Doubling