Return to model library
DeepSeek 2026-04-24
deepseek-v4-flash
DeepSeek V4 极速档 · 超低成本高并发 · 实时聊天
Universaltextultra low costExtremely fast reasoning
数据政策 / Data Policy:未知 / Unknown
未经法务确认上游 ToS,不做承诺
查看 DeepSeek 服务条款提示:此 provider 数据政策待法务审定;medical / legal 客户请联系商务确认 DPA。
context window
128K
tokens
maximum output
8.2K
tokens
knowledge cutoff
2025-12
Overall rating
7.57
/ 10
capability radar
code7.5
Mathematics7.0
reasoning7.5
creativity7.0
multilingual7.5
long context7.5
speed9.0
Pricing
input price$0.168000/ 1M tokens
output price$0.336000/ 1M tokens
Support features
Tool callJSON outputStreaming outputPrompt cache
Recommended scenarios
DeepSeek V4 极速档
超低成本高并发
实时聊天
Call example
Called via Nexevo.ai gateway — fully compatible with OpenAI SDK, just replace base URL
curl https://api.nexevo.ai/v1/chat/completions \
-H "Authorization: Bearer $NEXEVO_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "deepseek-v4-flash",
"messages": [
{ "role": "user", "content": "Hello!" }
]
}'