^{An unofficial free deepseek api wrapper written in python.}

A totally free deepseek api(with latest R1 model) written in python.

Required Cookies

Create an account in deepseek.com and visit chat.deepseek.com and copy the following cookies

smidV2
HMACCOUNT
Hm_lpvt_fb5acee01d9182aabb2b61eb816d24ff
intercom-session-guh50jw4
HWWAFSESTIME
Hm_lvt_fb5acee01d9182aabb2b61eb816d24ff
HWWAFSESID
.thumbcache_6b2e5483f9d858d7c661c5e276b6a6ae
Hm_lvt_1fff341d7a963a4043e858ef0e19a17c
intercom-device-id-guh50jw4
__cf_bm
ds_session_id
Hm_lpvt_1fff341d7a963a4043e858ef0e19a17c

and from Local Storage collect userToken value part

http headers

headers={

    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:134.0) Gecko/20100101 Firefox/134.0",
    "Accept": "*/*",
    "Accept-Language": "en-US,en;q=0.5",
    "Accept-Encoding": "gzip, deflate, br",
    "X-Client-Platform": "web",
    "X-Client-Version": "1.0.0-always",
    "X-Client-Locale": "en_US",
    "X-App-Version": "20241129.1",
    "Authorization": "Bearer {userToken}", #collected from local storage
    "Origin": "https://chat.deepseek.com", 
    "Referer": "https://chat.deepseek.com/", #change to 'https://chat.deepseek.com/a/chat/s/{chatroom_id}' after chatroom creation
    "Sec-Fetch-Dest": "empty",
    "Sec-Fetch-Mode": "cors",
    "Sec-Fetch-Site": "same-origin",
    "Priority": "u=0",
    "Te": "trailers",
}

Install packages

pip install -r requirements.txt

Usage

You can usage the terimnal wrapper this way

from deepseek import DeepSeek
import json
from rich.console import Console
from rich.markdown import Markdown
console = Console()


with open("cookies.json", "r") as f:
    cookies = json.load(f)
f.close()

usertoken = "....." #collect from locla storage

deepseek_engiene:DeepSeek = DeepSeek(cookie=cookies,user_token=usertoken)

Create a chatroom

chatroom_id = deepseek_engiene.create_chatroom()

#you can visit the chatroom https://chat.deepseek.com/a/chat/s/{chtroom_id}

Ai Response method blueprint

def ai_response(self,prompt:str,file_list:list=None,r1_enabled:bool=False,web_search:bool=False)->str:

    """
    Get response from the ai
    :param prompt: The prompt to send to the ai
    :param file_list: List of files to send to the ai
    :param r1_enabled: R1 enabled
    :param web_search: Web search enabled
    :return: The response from the ai
    """

Normal Ai Response

markdown = Markdown(deepseek_engiene.ai_response(prompt=prompt))
console.print(markdown)

Ai hallucination (R1-enabled)

markdown = Markdown(deepseek_engiene.ai_response(prompt=prompt,r1_enabled=True))
console.print(markdown)

Web search Enable

markdown = Markdown(deepseek_engiene.ai_response(prompt=prompt,web_search=True))
console.print(markdown)

File upload

markdown = Markdown(deepseek_engiene.ai_response(file_list=["assets/leetcode.png","assets/leetcode2.png"],prompt="solve the problems in python and complete the function"))

Example output

In the following example deepseek solves two leetcode question from screenshots

markdown = Markdown(deepseek_engiene.ai_response(file_list=["assets/leetcode.png","assets/leetcode2.png"],prompt="solve the problems in python and complete the function"))

CodeFlow (sorta)

1. Creating a chatroom

2. Util Function create_pow_challenge

3. Ai responses
   The api endpoint for this is /api/v0/chat/completion

   First Message

   Second Message

   Third message and onwards

   File upload

   Get file upload status

   Use the File id
   Use the file_id in the ref_file_ids array in the post request

Todo

1. Scrape for file upload
2. Code implementation in python.
3. chat history implementation
4. Start messaging in an already existing chatroom
5. Better error handling

Note

The cloudflare __cf_bm cookie expires every thirty minutes. So if the requests aren't working then refresh the cookies