EasyAPI

deepseek-v4.1-flash-1017

Text

DeepSeek-V4.1-Flash is DeepSeek's current mainstay model; its official API model name is deepseek-flash. It supports a 1M context and up to 384K output, with thinking and non-thinking modes switchable by default, features image understanding, tool use, and JSON Output, and is compatible with the OpenAI and Anthropic APIs.

Input price

¥0.4258% off

Market: ¥1

/ M

Output price

¥1.6858% off

Market: ¥4

/ M

Cache read

¥0.008458% off

Market: ¥0.02

/ M

Input

Text / Image

Output

Text

Peak / off-peak billing

Prices vary by time of day (UTC+8)

Currently off-peak

Time-of-day price table

  • Input
    Off-peak¥0.42
    Peak hours¥0.84
  • Output
    Off-peak¥1.68
    Peak hours¥3.36
  • Cache read
    Off-peak¥0.0084
    Peak hours¥0.0168
FieldOff-peakPeak hours
Input¥0.42¥0.84
Output¥1.68¥3.36
Cache read¥0.0084¥0.0168

Peak hours: UTC+8 Mon–Fri (excluding Chinese statutory holidays) 09:00 - 12:00、14:00 - 18:00; all other periods are off-peak. The billing period is based on UTC+8 time when the request completes

API

Quick Start

Drop-in code to call this model with EasyAPI's OpenAI-compatible API.

1

Get your API key

Create an API key from your dashboard and set it as an environment variable:

Replace sk-xxxxxxxx with your API token from the console (Bearer token).

Create API token
shell
export EASYAPI_API_KEY=sk-xxxxxxxx
2

Make your first request

Use deepseek/deepseek-v4.1-flash-1017 with the EasyAPI API:

cURL
curl https://token.easyapi.com/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $EASYAPI_API_KEY" \
  -d '{
  "model": "deepseek/deepseek-v4.1-flash-1017",
  "messages": [
    {
      "role": "user",
      "content": "Hello!"
    }
  ]
}'
3

Enable streaming

Add "stream": true to your request body to receive responses as server-sent events:

curl
curl -N https://token.easyapi.com/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $EASYAPI_API_KEY" \
  -d '{"model":"deepseek/deepseek-v4.1-flash-1017","messages":[{"role":"user","content":"Hello!"}],"stream":true}'

Endpoint

Sends a chat completion request. Supports streaming and non-streaming modes.

POSThttps://token.easyapi.com/v1/responses
Authorization
Bearer $EASYAPI_API_KEY
Content-Type
application/json
Model
deepseek/deepseek-v4.1-flash-1017

Send a chat completion request, with streaming or non-streaming responses. Docs

GEThttps://token.easyapi.com/v1/models
Authorization
Bearer $EASYAPI_API_KEY

List the models available to the current account. Docs

Parameters

NameTypeDefaultDescription
modelstring—Model ID, e.g. openai/gpt-4o or anthropic/claude-3-5-sonnet
messagesarray—List of conversation messages, each with role and content
streambooleanfalseWhen true, the response is streamed via SSE
temperaturefloat1Sampling temperature; higher values are more random
max_tokensinteger—Maximum number of tokens generated in a single reply
top_pfloat1Nucleus sampling; limits the cumulative probability of candidate tokens
frequency_penaltyfloat0Frequency penalty; reduces repeated wording
presence_penaltyfloat0Presence penalty; encourages new topics

Using third-party SDKs

Official OpenAI and other SDKs work by swapping the base URL. See integration guide

deepseek-v4.1-flash-1017 FAQ

How is the deepseek-v4.1-flash-1017 API billed?
deepseek-v4.1-flash-1017 is pay-as-you-go from ¥0.42 / M. The exact rate is listed in the pricing table on this page and billed at EasyAPI's public price, with no subscription required.
How do I call deepseek-v4.1-flash-1017 through the API?
EasyAPI exposes an OpenAI-compatible API: create an API key in the console, set base_url to https://token.easyapi.com and model to deepseek/deepseek-v4.1-flash-1017, then send requests to /v1/responses. The Quick Start on this page includes cURL, Python and TypeScript examples; official OpenAI SDKs only need the base_url swapped.
Which request formats does deepseek-v4.1-flash-1017 support?
OpenAI Chat Completions format including streaming (stream). Official OpenAI and Anthropic SDKs work by swapping base_url; see the Parameters table on this page for the full list.
How do I get started with deepseek-v4.1-flash-1017?
Sign up for EasyAPI, create an API key in the console and top up your balance to call deepseek-v4.1-flash-1017 pay-as-you-go, no subscription needed. You can also try it in the online Playground first, then copy the Quick Start example.

Related models

All models

deepseek-v4-pro-0813

DeepSeek

The DeepSeek V4 flagship, with the series' strongest deep reasoning and code generation, suited to whole-codebase analysis, multi-step automation and demanding instructions.

Text
55% off
Input price

¥1.981/ M

Output price

¥5.943/ M

Cache read

¥0.066/ M

Cache write

—

doubao-seed-2.0-lite

ByteDance

Lightweight tier of the Seed 2.0 family tuned for low-latency agent, coding, and GUI workloads.

Text
6% off
Input price

¥0.5557/ M

Output price

¥3.334/ M

Cache read

¥0.1111/ M

Cache write

—

doubao-seed-2.0-mini

ByteDance

Lightweight tier in the Seed 2.0 family, tuned for high-throughput, low-latency tasks with adjustable reasoning levels. Fits batch processing, moderation, and classification.

Text
6% off
Input price

¥0.1852/ M

Output price

¥1.8522/ M

Cache read

¥0.037/ M

Cache write

—

doubao-seed-2.0-pro

ByteDance

Flagship of the Seed 2.0 series for complex reasoning and long-horizon agent workflows with reliable tool use.

Text
6% off
Input price

¥2.9636/ M

Output price

¥14.8178/ M

Cache read

¥0.5927/ M

Cache write

—

Popular comparisons for deepseek-v4.1-flash-1017