apistore.cloud
Log inGet API key
DeepSeek V4.1 · OpenAI-compatible

DeepSeek API access with one key and prepaid billing

Call DeepSeek V4.1 for chat, coding and agents through an OpenAI-compatible API. Change the base URL and the model name, keep the rest of your code, and pay only for the tokens you use.

Get API keyRead the docs
1M tokens of contextUp to 393.2K output tokensStreaming and tool callsCheaper off-peak
A BytePlus official partner

Price

US$ per 1M tokens. The first figure is what you pay; the list price is struck through. Off-peak prices apply outside the peak hours shown under each model.

ModelContextInput (cache hit)Input (cache miss)Output
deepseek-v4-1-flash
Peak hours: Mon–Fri 09:00–12:00, 14:00–18:00 UTC+8
−30% off the list price		1M
$0.0042 $0.006 · peak
$0.0021 $0.003 · off-peak
$0.21 $0.30 · peak
$0.105 $0.15 · off-peak
$0.84 $1.20 · peak
$0.42 $0.60 · off-peak
Try it
deepseek-v4-1-flash-economy
Peak hours: Mon–Fri 09:00–12:00, 14:00–18:00 UTC+8
−30% off the list price		1M
$0.0035 $0.005 · peak
$0.0014 $0.002 · off-peak
$0.168 $0.24 · peak
$0.084 $0.12 · off-peak
$0.672 $0.96 · peak
$0.336 $0.48 · off-peak
Try it

See all prices →

Copy, paste, run

The same request you send to any OpenAI-compatible API. Works with the OpenAI SDKs, LangChain, Cursor, Cline, Claude Code and any MCP client.

from openai import OpenAI

client = OpenAI(base_url="https://api.apistore.cloud/v1", api_key="sk-...")

stream = client.chat.completions.create(
    model="deepseek-v4-1-flash",
    messages=[{"role": "user", "content": "Hello!"}],
    stream=True,
)
for chunk in stream:
    print(chunk.choices[0].delta.content or "", end="")

Setup guides: Cursor · Cline · Claude Code

How you pay

  • Prepaid balance with no subscription and no monthly minimum. Top up by card from US$20; your balance does not expire.
  • Every request is listed with its tokens and cost. Set a daily spending cap on each key.
  • Requests that fail before an answer are not charged.

Limits and reliability

  • Context window of 1M tokens, up to 393.2K output tokens per answer.
  • Limits apply per key (requests per minute) and per account (concurrent requests).
  • If a provider is slow or down, your request moves to another one automatically.

Questions

Which DeepSeek models can I call?

Today: deepseek-v4-1-flash, deepseek-v4-1-flash-economy. The pricing page always lists what is on sale.

How do I get a DeepSeek API key?

Create an account with email or Google, verify your address, then create a key on the API keys page. Add credit by card and the key works at once.

How am I billed?

Per token, from your prepaid balance: input tokens (cheaper when they are read from cache) and output tokens, at the peak or off-peak price of the moment. Nothing is charged for a request that fails before an answer.

Does it work with Cursor, Cline and Claude Code?

Yes. Any tool that accepts an OpenAI-compatible base URL and key works. There is also a hosted MCP server and the apistore command-line tool.

Start in a few minutes

Create an account, verify your email, create a key and make your first call.

Get API keySee all prices