Third-party

DeepSeek V4 Flash

Lightweight DeepSeek MoE model for fast, low-latency, cost-effective everyday dialogue, content, RAG, and batch work.

Request quote review Compare prices

Model details

Model code

deepseek-v4-flash

Input $0.138 / 1M tokens · Output $0.275 / 1M tokens

Input

$0.138 / 1M tokens

Input(Implicit Cache)

$0.028 / 1M tokens

Output

$0.275 / 1M tokens

Detail checked

Source region: Chinese Mainland. Current highlighted entries were rechecked on July 21, 2026; long-tail entries retain their earlier official detail evidence where applicable. Final quotes still require official console confirmation for region, account route, quota, promotions, taxes, and current availability.

Buyer review

Questions to confirm before purchase

Does this exact model code support the buyer's region?

Is this for Token Plan, direct API, or both?

Does the workload need text, image, video, audio, or embeddings?

Are context length, rate limits, and quota enough for production?

Are official usage cost and ModelSmarter service fee separated?

Is there a lower-cost fallback model if usage grows?

Source note

Current highlighted model coverage was checked against the official Alibaba Cloud Model Studio model page on 2026-07-21. Existing long-tail detail summaries retain their earlier console evidence date where noted. Availability, region, account route, quota, taxes, promotions, and official terms must be confirmed before purchase.

Open official console source