منتشرشده در 2026-08-16 · 赵予安
پاسخ مستقیم
DeepSeek V4 Flash CN is listed at 0.020 / 0.041 per million tokens for Chinese workloads. Context window and cache safety stay unpublished. این راهنما برای تیمهای محصول و پلتفرم نوشته شده که کیفیت مدل، هزینه، سیاست مسیریابی و ریسک انتشار را مقایسه میکنند.
What shipped
Deepseek V4 Flash CN is listed on the NextModel gateway as deepseek/deepseek-v4-flash-cn. The public slug is deepseek--deepseek-v4-flash-cn. The provider is DeepSeek. The model has streaming capability. There is no first-party CacheSafety row for this exact model in the NextModel bench table, so Safe Hit, Bad Hit, and trap rates are unpublished.
Pricing and context
Input price is $0.020 per 1M tokens. Output price is $0.041 per 1M tokens. The context window is unpublished. The latency band is 1000-3000ms. For comparison, Deepseek V3.2 CN is $0.042 per 1M input tokens and $0.064 per 1M output tokens. Deepseek V4 PRO CN is $0.239 per 1M input tokens and $0.479 per 1M output tokens.
Who it is for
The model is for Chinese-language use cases. Its price sits below both Deepseek V3.2 CN and Deepseek V4 PRO CN. Streaming output makes it usable for incremental response delivery.
What is still unverified
The release date is unpublished in this catalog snapshot. Independent third-party quality scores are not in these notes. The context window is unpublished when the catalog value is 0. CacheSafety rates for this exact model are unavailable.