deepseek-v4-flash-0731

DeepSeekTime-of-day pricing

DeepSeek V4 Flash 0731 is a sparse mixture-of-experts model from DeepSeek, with 13B active parameters out of 284B total. This re-post-trained revision is the GA release of DeepSeek V4 Flash, suited for coding, reasoning, and agent workflows. It has a 1,048,576-token context window, accepts text only, and supports thinking mode, tool calling, and structured outputs. DeepSeek V4 Flash 0731 是 DeepSeek V4 Flash 的正式版,稀疏混合专家模型,总参数 2840 亿、激活 130 亿。面向编程、推理和智能体工作流。上下文 100 万 token,只接受文本,支持思考模式、工具调用和结构化输出。

  • 学术
  • 金融
  • 医疗
  • 法律
  • 营销
  • 搜索优化
  • 编程
  • 角色扮演
  • 科学
  • 技术
  • 翻译
  • 百科

Quick start

curl https://llmoxy.com/v1/chat/completions \  -H "Authorization: Bearer $LLMOXY_KEY" \  -H "Content-Type: application/json" \  -d '{    "model": "deepseek-v4-flash-0731",    "messages": [{"role": "user", "content": "Hello"}]  }'

Set the LLMOXY_KEY environment variable to your API key before running.

Endpoints

  • Chat CompletionsPOST /v1/chat/completions

FAQ

How much does deepseek-v4-flash-0731 cost on LLMoxy?
Time-of-day pricing. Price in the “Deepseek” token group: From In ¥0.06 / 1M. That is 94% below the official API price.
How do I call deepseek-v4-flash-0731?
Create an API key in the LLMoxy console, set model to deepseek-v4-flash-0731, and send the request to: Chat Completions POST https://llmoxy.com/v1/chat/completions. You can also use the OpenAI SDK directly with the base URL set to https://llmoxy.com/v1.
Which token groups can I use with deepseek-v4-flash-0731?
The token group sets the price you pay: you pick a group when you create an API key in the console, and every call made with that key is billed at that group’s price. Groups available for deepseek-v4-flash-0731: Deepseek.
Do I need an international credit card?
No. Top up your balance in the console first; each call is then deducted from the balance based on actual usage. Payment methods: Alipay, WeChat Pay, Visa / MasterCard, Apple Pay, and crypto such as USDT, BTC, SOL, and TRX. Multi-currency support.
All Models