ModelsDeepSeek-V4.1-Flash

DeepSeek-V4.1-Flash

deepseek-flashDeepSeek

Token

DeepSeek-V4.1-Flash；官方调用名 deepseek-flash。文本/图像输入，文本输出；支持思考与非思考、JSON Output、工具调用。刊例价为人民币/百万 tokens，按北京时间工作日 09:00–12:00、14:00–18:00 高峰及其余空闲时段计价。来源：https://api-docs.deepseek.com/zh-cn/quick_start/pricing/

Quick Start Full API Documentation

Context Window

1,000,000

Max Output

384,000

Quality

—

Supply Type

Official

Pricing

Input

¥1.00

per 1M tokens

•

Output

¥4.00

per 1M tokens

•

Cache

¥0.02

per 1M tokens

Pricing Details

Billing item	Unit	Applies when	List price
Cache read cache_read_tokens	per 1M tokens	Off peakTimezone: Asia/Shanghai	¥0.02CNY
Input tokens input_tokens	per 1M tokens	Off peakTimezone: Asia/Shanghai	¥1.00CNY
Output tokens output_tokens	per 1M tokens	Off peakTimezone: Asia/Shanghai	¥4.00CNY
Cache read cache_read_tokens	per 1M tokens	PeakTimezone: Asia/Shanghai	¥0.04CNY
Input tokens input_tokens	per 1M tokens	PeakTimezone: Asia/Shanghai	¥2.00CNY
Output tokens output_tokens	per 1M tokens	PeakTimezone: Asia/Shanghai	¥8.00CNY

Capabilities

Code generation Reasoning Text generation Tool use Visual understanding

Input:Text, Image

Output:Text

Model Specs

Context Window1,000,000

Max Output384,000

Pricing KindToken

Supply TypeOfficial

FamilyDeepSeek

Quick Start

Set ROUTERSHIFT_API_KEY in your environment before running this example. Copying code does not send a request.

curl "https://api.routershift.com/v1/chat/completions" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $ROUTERSHIFT_API_KEY" \
  -d '{
  "model": "deepseek-flash",
  "messages": [
    {
      "role": "user",
      "content": "Hello!"
    }
  ],
  "max_tokens": 100
}'

Full API Documentation

Example parameters

3 example presets, not community statistics. Check model support before use.

Balanced

temp0.7

top_p0.9

max_tokens4,096

Creative

temp0.9

top_p1.0

max_tokens8,192

Precise

temp0.3

top_p0.8

max_tokens2,048