ModelSmarterStart intake
ModelsPlansDocs
Back to model catalog

Third-party

DeepSeek V4 Flash

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

Request quote reviewCompare prices

Model details

Model code

deepseek-v4-flash

Category

Third-party

Family

DeepSeek

Capability

Fast reasoning

Modality

Text -> Text

Release / status

2026-05-11

Snapshot

Current model code

Source region

Chinese Mainland

Official detail price

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. This is a copied summary from the official Model Studio detail page checked on June 8, 2026. 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

Catalog taxonomy and model detail price summaries were checked against Alibaba Cloud Model Studio Console on 2026-06-08. Availability, region, account route, quota, taxes, promotions, and official terms must be confirmed before purchase.

Open official console source