Issue is still there Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'.

[ranjan-sumit]

Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'.****

[dsdanielpark]

Not package-related issue, but possible action on the account due to unauthorized usage (excessive traffic).

I have confirmed that there are no major issues with the current functioning of the package, and it is operating normally.

As mentioned in various issues and the README.md, bardapi was implemented to support developers in mocking and implementing certain testing functionalities before the official package distribution of bard. Please note that the token values used are associated with Google accounts.

This particular issue could be a precautionary measure against unauthorized usage (such as excessive traffic), and if you continue to use it, there is a possibility of encountering issues with your Google account.

Therefore, I strongly advise against using this package to generate excessive traffic and recommend checking the official API documentation instead.

97 #99

Therefore, various types of errors (Nonetype, SNIM0e value, Response Error) may occur, but these are not package-related issues and could be actions taken by Google regarding the account.

We recommend trying with the cookie value of a different account or checking the functionality in Google Colab or Logout and Re-Login. According to other users, it should return to normal operation after a few hours. Thank you.

[nikk-16]

There is no issue int this API. But after excessive usage, the Google knows that cookie is having unauthorized access so the 1PSID is revert in some form.

The reason we get : Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'

How can you get rid of it: Clear cookies of bard.google.com and again

• login into the account
• F12 > Applications > Cookies > bard > __Secure-1PSID
• Copy cookie and paste into your code.

Re run and you are good to go.

[pixkk]

Admin. We need use two cookies - Secure-1PSID and Secure-1PSIDTS. Without __Secure-1PSIDTS html page bard.google.com has unauthorized user.

[dsdanielpark]

If you need to set multi cookie value. Using Bard Cookies Object.

from bardapi import BardCookies

cookie_dict = {
    "__Secure-1PSID": "xxxxxxxxx",
    "__Secure-1PSIDTS": "xxxxxxxxx",
    # Any cookie values you want to pass session object.
}

bard = BardCookies(cookie_dict=cookie_dict)

print(bard.get_answer("こんにちは"))

[TanukiAI]

On Windows, importing BardCookies is not possible:

>>> from bardapi import BardCookies
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: cannot import name 'BardCookies' from 'bardapi' (C:\Program Files\Python310\lib\site-packages\bardapi\__init__.py)

[dsdanielpark]

Upon rechecking, I confirmed that there were no issues with the implementation of BardCookies.

Additionally, I remembered that it was separated as a standalone entity to avoid any potential issues arising in the Bard class. Therefore, your suggestion was already present in BardCookies, and it was functioning properly on both Windows and Linux.

Check version and you can use BardCookies that exatly same except can pass cookie dictionary as argument.

from bardapi import BardCookies

cookie_dict = {
    "__Secure-1PSID": "xxxxxxxxx",
    "__Secure-1PSIDTS": "xxxxxxxxx",
    "__Secure-1PSIDCC", "xxxxxxx")
}

bard = BardCookies(cookie_dict=cookie_dict)

print(bard.get_answer("こんにちは"))

[TanukiAI]

But there is a workaround for this looking at the code:

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Wouldn't it be possible to pass a dictionary with cookies along?

[TanukiAI]

Here in Austria, it doesn't work with just the one token. And even if it's just a minority, I don't think there would be a problem passing an cookie dictionary (basically {"cookie1": "value1", "cookie2": "value"}) along and setting these cookies, too. Creating the session with the multiple cookies worked.

Unfortunately I can't help currently as I don't have an IDE here and can't install one, I am trying all of this on Windows in Notepad++ plus CLI :/

[parsibox]

But there is a workaround for this looking at the code:

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Wouldn't it be possible to pass a dictionary with cookies along?

this work for me thanks

[dsdanielpark]

@TanukiAI

On Windows, importing BardCookies is not possible:

>>> from bardapi import BardCookies
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: cannot import name 'BardCookies' from 'bardapi' (C:\Program Files\Python310\lib\site-packages\bardapi\__init__.py)

You can still use the Bard class as it is and also have the option to use BardCookies. In fact, both BardCookies and Bard have nearly identical implementations, except that BardCookies can accept additional cookie values as a dictionary.

However, due to my current commitments to another project, I won't be able to address any additional issues that may arise. Therefore, I have decided to maintain this version as it is for now. The development priority has been postponed until a later time when I can address any issues and focus on adding new features.

Anyone is welcome to contribute at any time. Thank you for your understanding.

How to use BardCookies

BardCookies works fine without any differences on both Windows and Linux. Please reinstall it to ensure you have the latest version.

1. Reinstall Lastest Version of Bard-API

   pip install bardapi>=0.1.24

2. And check version

   import bardapi
   bardapi.__version__
   >>> '0.1.24'

3. Using BardCookies

   
   from bardapi import BardCookies

cookie_dict = { "Secure-1PSID": "xxxxxxxxx", "Secure-1PSIDTS": "xxxxxxxxx", "__Secure-1PSIDCC", "xxxxxxx") }

bard = BardCookies(cookie_dict=cookie_dict)

print(bard.get_answer("こんにちは"))

For more feature of Bard-API package, please check this [document](https://github.com/dsdanielpark/Bard-API/blob/main/README_DEV.md). 

[noseon]

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard
import json
import os

def load_cookie(element_name):
    with open(os.getcwd() + "/" + "bard.json", "r", encoding="utf-8") as file:
        data = json.load(file)
        if isinstance(data, list):
            for item in data:
                if item.get("name") == element_name:
                    return item.get("value")
        return None

_1PSID = load_cookie("__Secure-1PSID")
_1PSIDTS = load_cookie("__Secure-1PSIDTS")
_1PSIDCC = load_cookie("__Secure-1PSIDCC")
session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", _1PSID)
session.cookies.set("__Secure-1PSIDTS", _1PSIDTS)
session.cookies.set("__Secure-1PSIDCC", _1PSIDCC)

bard = Bard(token=_1PSID, session=session)
print(bard.get_answer("gnomes")['content'])

this working if put in the heart of the api would be better. just passing the json location to get the data

[nerblock]

Same problem when trying to access Bard from Hungary. __Secure-1PSIDTS needs to be included in the request, and it needs to be rotated every ~20minutes. In the chromium source it is called a bound session cookie:

https://github.com/nwjs/chromium.src/blob/21945e5c163533b01a44214b663a9bf615c31ab5/chrome/browser/signin/bound_session_credentials/bound_session_cookie_refresh_service_impl.cc#L248

[noseon]

so in this case it would be reloading each question now the section just use __Secure-1PSID?

[TanukiAI]

Yes, cookies need to be rotated. An request is sent to https://accounts.google.com/RotateCookies to renew the token.

[nerblock]

Ok, so the million dollar question is: can we do this without a browser? Because having to replace this cookie manually every 10-20 minutes is bad news for my use case.

[dsdanielpark]

Ok, so the million dollar question is: can we do this without a browser? Because having to replace this cookie manually every 10-20 minutes is bad news for my use case.

While support for this issue is available from version 0.1.26 onwards, it may still be challenging to find a definitive solution. https://github.com/dsdanielpark/Bard-API#auto-cookie-bard

Auto Cookie Bard

Using browser_cookie3 we extract the `__Secure-1PSID`` cookie from all browsers, and then we can use the API without passing the token. However, there are still incomplete dependency packages and various variables, so please seek assistance in the following GitHub Issues or adjust your browser's version.

from bardapi import Bard

bard = Bard(token_from_browser=True)
res = bard.get_answer("Do you like cookies?")
print(res)

• Release Note: https://github.com/dsdanielpark/Bard-API/releases/tag/0.1.26

[rajserc]

Going to Incognito and collecting the cookie (from the new signin inside incognito) seems to help.

[znishan]

I agree with @rajserc. The issue seems to be related to being signed to multiple Google accounts in the same browser session. Opening an Incognito window, and logging into Bard to capture__Secure-1PSID worked for me.

I suppose it's worth a try to log out of all but one account before capturing the Bard token.

[noseon]

browser_cookie3 using the firefox version it seems that you have to get the sessionstore.js file if that's the case you could make the bard api read this file and that's it it will work continuously and lastingly

[doffn]

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard
import json
import os

def load_cookie(element_name):
    with open(os.getcwd() + "/" + "bard.json", "r", encoding="utf-8") as file:
        data = json.load(file)
        if isinstance(data, list):
            for item in data:
                if item.get("name") == element_name:
                    return item.get("value")
        return None

_1PSID = load_cookie("__Secure-1PSID")
_1PSIDTS = load_cookie("__Secure-1PSIDTS")
_1PSIDCC = load_cookie("__Secure-1PSIDCC")
session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", _1PSID)
session.cookies.set("__Secure-1PSIDTS", _1PSIDTS)
session.cookies.set("__Secure-1PSIDCC", _1PSIDCC)

bard = Bard(token=_1PSID, session=session)
print(bard.get_answer("gnomes")['content'])

this working if put in the heart of the api would be better. just passing the json location to get the data

it worked for me with out using json

[chandrakalagowda]

There is no issue int this API. But after excessive usage, the Google knows that cookie is having unauthorized access so the 1PSID is revert in some form.

The reason we get : Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'

How can you get rid of it: Clear cookies of bard.google.com and again

• login into the account
• F12 > Applications > Cookies > bard > __Secure-1PSID
• Copy cookie and paste into your code.

Re run and you are good to go.

This worked for me. Thank you.

I am pasting the code for beginners like me who might find it helpful

[Stainless5792]

But there is a workaround for this looking at the code:

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Wouldn't it be possible to pass a dictionary with cookies along?

works for me!

[OleksandrCherniavskyi]

Run browser in incognito version and try to get token, it's work for me.

[Aulig]

Yes, cookies need to be rotated. An request is sent to https://accounts.google.com/RotateCookies to renew the token.

Did you find out how that request works? @TanukiAI

[Bharat9114]

Not package-related issue, but possible action on the account due to unauthorized usage (excessive traffic).

I have confirmed that there are no major issues with the current functioning of the package, and it is operating normally.

As mentioned in various issues and the README.md, bardapi was implemented to support developers in mocking and implementing certain testing functionalities before the official package distribution of bard. Please note that the token values used are associated with Google accounts.

This particular issue could be a precautionary measure against unauthorized usage (such as excessive traffic), and if you continue to use it, there is a possibility of encountering issues with your Google account.

Therefore, I strongly advise against using this package to generate excessive traffic and recommend checking the official API documentation instead.

97 #99

Therefore, various types of errors (Nonetype, SNIM0e value, Response Error) may occur, but these are not package-related issues and could be actions taken by Google regarding the account.

We recommend trying with the cookie value of a different account or checking the functionality in Google Colab or Logout and Re-Login. According to other users, it should return to normal operation after a few hours. Thank you.

Hey Daniel , can we use secure 2PSID or secure 3PSID for this is there another way of re engineering that this will use as api for success . Like i used 2-3 query only it got stopped .

[mchappychen]

Auto Cookie Bard

Hello, does this require a browser to be open and running?

Unfortunately I work on a linux server that doesnt have any browsers.

I was wondering if someone else had the idea of visiting bard.google.com using the requests module like so:

import requests
url = 'https://bard.google.com/'
cookies = {
    '__Secure-1PSID': '--',
    '__Secure-1PSIDCC':'--',
    '__Secure-1PSIDTS':'--'
}
r = requests.get(url, cookies=cookies)
print(r.cookies._cookies['.google.com']['/'])

I was able to login successfully using some of the cookies listed in that example code, but I wasn't able to retrieve all cookies. I'm guessing it's because some cookies are httpOnly type, and those are harder to get.

Update: After some testing, I'm able to call bard with just a Token, and it never expires. I think this is because after calling bard from python, I can no longer visit the bard.google.com website or else it'll reset my token.

[ChrysthianChrisley]

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Works fine!

[mohan-chinnappan-n]

Here is chrome extension to get Cookies : https://chrome.google.com/webstore/detail/cookie-viewer/dedhcncdjkmjpebfohadfeeaopiponca

[zoftdev]

On Windows, importing BardCookies is not possible:

>>> from bardapi import BardCookies
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: cannot import name 'BardCookies' from 'bardapi' (C:\Program Files\Python310\lib\site-packages\bardapi\__init__.py)

please use latest version

[naveedazam27]

Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'.****

clear browser cookies and then try with new cookies data it worked for me :D

[voaneves]

There is no issue int this API. But after excessive usage, the Google knows that cookie is having unauthorized access so the 1PSID is revert in some form.

The reason we get : Exception: SNlM0e value not found. Double-check __Secure-1PSID value or pass it as token='xxxxx'

How can you get rid of it: Clear cookies of bard.google.com and again

• login into the account
• F12 > Applications > Cookies > bard > __Secure-1PSID
• Copy cookie and paste into your code.

Re run and you are good to go.

It's my first time using it, but your solution actually fixed it for me.

Thanks!

[Johan0401]

Hello all, Hope you do well. I have tried to make an AI chatbot that can talk to me instead of shooting out words. But it gives me all kinds of errors. I no nothing about coding i Just started out. So here is my code i would appreciate it if someone can tell me what i did wrong. `!pip install bardapi from bardapi import Bard import os import time os.environ['_BARD_API_TOKEN'] = "Insert.API here"

Set your input text

input_text = "Why is the sky blue?"

input_text = "What is the current status of Ashes Series 2023" print(Bard().get_answer(input_text)['content'])` It gives me the following error. Exception Traceback (most recent call last) " style="color: var(--colab-anchor-color);"> in <cell line: 5>() 3 input_text = "Why is the sky blue?" 4 # input_text = "What is the current status of Ashes Series 2023" ----> 5 print(Bard().get_answer(input_text)['content'])

1 frames

/usr/local/lib/python3.10/dist-packages/bardapi/core.py in _get_snim0e(self) 114 """ 115 if not self.token or self.token[-1] != ".": --> 116 raise Exception( 117 "Secure-1PSID value must end with a single dot. Enter correct Secure-1PSID value." 118 ) Exception: Secure-1PSID value must end with a single dot. Enter correct Secure-1PSID value. The API has a dot, so i don't know why it is not working 😕 would appreciate the help

[dsdanielpark]

Is this code work?

token = "Insert.API here"

input_text = "What is the current status of Ashes Series 2023"
print(Bard(token).get_answer(input_text)['content'])`

[Johan0401]

I know I gave me the error message I have said above. It says something about the API can't be Authorised.

[Kienvu175]

I just simply clear cookies, refresh bard,google.com then re-copied the 1PSID then it worked. Hope this help !

[Ayad-Mihidabi-Khan-Jitu]

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Works fine!

THANKS WORKED

[vincecalpari]

Same problem when trying to access Bard from Hungary. __Secure-1PSIDTS needs to be included in the request, and it needs to be rotated every ~20minutes. In the chromium source it is called a bound session cookie:

https://github.com/nwjs/chromium.src/blob/21945e5c163533b01a44214b663a9bf615c31ab5/chrome/browser/signin/bound_session_credentials/bound_session_cookie_refresh_service_impl.cc#L248

Tested and verified above, same issue in Sri Lanka, the __Secure-1PSIDTS rotates every ~20mins. Secure-1PSID & Secure-1PSIDCC can be kept the same, but we have to manually retrieve the Secure-1PSIDTS key every 20 mins. Any work around for this concern? Using bard api as a learning task, help would be appreciated!

session.cookies.set("__Secure-1PSID", "same-xxx") session.cookies.set( "__Secure-1PSIDCC", "same-xxx") session.cookies.set("__Secure-1PSIDTS", "change-every-20-mins")

[lionkingsiro]

how to refresh and get new Secure-1PSIDTS or Secure-3PSIDTS by java code? i block all xhr request with out https://mobilesdk-pa.clients6.google.com/v1/projects/xxxx:queryBilledUsage, and browser try create new token and get succes. I think __Secure-3PSIDTS generated by javascript.

[phamxtien]

I fix it by update cookie after every request cookies = [] for k, v in bard.session.cookies.get_dict().items(): cookies.append({'name': k, 'value': v})

[digitalboy]

But there is a workaround for this looking at the code:

import requests
from bardapi.constants import SESSION_HEADERS
from bardapi import Bard

token = "xxxxx"

session = requests.Session()
session.headers = SESSION_HEADERS
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", "<VALUE>")
session.cookies.set("__Secure-1PSIDCC", "<VALUE>")

bard = Bard(token=token, session=session)

Wouldn't it be possible to pass a dictionary with cookies along?

Thank you! It is working.........

[shaggy2626]

I've been experimenting with the Bard API for some time now, and I seem to be encountering a couple of problems that I hope someone can help me clarify.

The main objective is to upload images via the Bard API. While I have managed to get this working, I've noticed that the tokens refresh frequently. When processing a batch of 50-100 images, the tokens tend to change after the first 5 responses.

I've been following the discussion in this thread and tried implementing browser_cookie3 as suggested in the Bard API's home page. However, while my script runs fine, it's not practical to update the tokens manually each time. I was wondering if anyone has been able to successfully use browser_cookie3 to pull all three parameters from the Bard cookies: __Secure-1PSIDTS, __Secure-1PSIDCC, __Secure-1PSID.

While I have access to Visual Question Answering (VQA), I find the Bard UI more user-friendly. The prompt length restriction (80 tokens max) in VQA also makes Bard a more preferable option.

If anyone has managed to resolve the token refresh issue without having to manually update the tokens each time, I would greatly appreciate your insights.

Here's the current script that I'm using:

import requests
from bardapi import Bard, SESSION_HEADERS
from PIL import Image
import pandas as pd
import io

def main():
    session = requests.Session()
    token = "insert here"

    # Set your cookies
    session.cookies.set("__Secure-1PSID", token)
    session.cookies.set("__Secure-1PSIDCC", "insert here")
    session.cookies.set("__Secure-1PSIDTS", "insert here")

    # Set the session headers
    session.headers = SESSION_HEADERS

    # Send an API request and get a response
    bard = Bard(token=token, session=session)

    # Load the text file and read the URLs
    with open('C:\\Users\\Downloads\\SectionalShape.txt', 'r') as f:
        urls = f.readlines()

    # Iterate over the URLs
    for url in urls:
        url = url.strip()  # Remove leading/trailing whitespace (including newline)

        # Get the image from the URL
        response = requests.get(url)

        # Debug: Print the status code and content type of the response
        print(f"Status code: {response.status_code}")
        print(f"Content-Type: {response.headers['content-type']}")

        try:
            # Open the image and convert it to bytes
            image = Image.open(io.BytesIO(response.content))
            byte_arr = io.BytesIO()
            image.save(byte_arr, format='JPEG')
            image_data = byte_arr.getvalue()
        except UnidentifiedImageError:
            print(f"Unable to open image from URL: {url}")
            continue

        # Ask the Bard API about the image
        bard_answer = bard.ask_about_image('What is in the image?', image_data)

        # Write the URL and response to the text file
        with open('C:\\Users\\Downloads\\SectionalShape.txt', 'a') as f:
            f.write(f"{url},{bard_answer['content']}\n")

if __name__ == "__main__":
    main()

Thank you in advance!

---

[nguyenduclong1147]

I created function get cookies from chorme. You can fix the error PermissionError: [Errno 13] Permission denied at https://github.com/borisbabic/browser_cookie3/issues/180#issuecomment-1587691954

  def _extract_bard_cookie() -> dict:
        cookie_dict = {}

        cj = browser_cookie3.chrome(cookie_file=r'C:\Users\<your user>\AppData\Local\Google\Chrome\User Data\Default\Network\Cookies',domain_name=".google.com")

        for cookie in cj:
            if cookie.name == "__Secure-1PSID" and cookie.value.endswith("."):
                cookie_dict["__Secure-1PSID"] = cookie.value
            if cookie.name == "__Secure-1PSIDTS":
                cookie_dict["__Secure-1PSIDTS"] = cookie.value
            if cookie.name == "__Secure-1PSIDCC":
                cookie_dict["__Secure-1PSIDCC"] = cookie.value

        logging.info(cookie_dict)
        return cookie_dict

[shaggy2626]

I created function get cookies from chorme. You can fix the error PermissionError: [Errno 13] Permission denied at borisbabic/browser_cookie3#180 (comment)

  def _extract_bard_cookie() -> dict:
        cookie_dict = {}

        cj = browser_cookie3.chrome(cookie_file=r'C:\Users\<your user>\AppData\Local\Google\Chrome\User Data\Default\Network\Cookies',domain_name=".google.com")

        for cookie in cj:
            if cookie.name == "__Secure-1PSID" and cookie.value.endswith("."):
                cookie_dict["__Secure-1PSID"] = cookie.value
            if cookie.name == "__Secure-1PSIDTS":
                cookie_dict["__Secure-1PSIDTS"] = cookie.value
            if cookie.name == "__Secure-1PSIDCC":
                cookie_dict["__Secure-1PSIDCC"] = cookie.value

        logging.info(cookie_dict)
        return cookie_dict

I tried that but im getting permission denied is this still working for you?

Traceback (most recent call last):
  File "C:\barderlatestgreat\bard_test.py", line 113, in <module>
    main()
  File "C:\barderlatestgreat\bard_test.py", line 26, in main
    cookie_dict = _extract_bard_cookie()
                  ^^^^^^^^^^^^^^^^^^^^^^
  File "C:\barderlatestgreat\bard_test.py", line 12, in _extract_bard_cookie
    cj = browser_cookie3.chrome(cookie_file='C:\\Users\\abc40\AppData\\Local\\Google\\Chrome\\User Data\\Default\\Network\\Cookies', domain_name=".google.com")
         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\site-packages\browser_cookie3\__init__.py", line 1160, in chrome
    return Chrome(cookie_file, domain_name, key_file).load()
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\site-packages\browser_cookie3\__init__.py", line 489, in load
    with _DatabaseConnetion(self.cookie_file) as con:
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\site-packages\browser_cookie3\__init__.py", line 349, in __enter__
    return self.get_connection()
           ^^^^^^^^^^^^^^^^^^^^^
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\site-packages\browser_cookie3\__init__.py", line 383, in get_connection
    con = method()
          ^^^^^^^^
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\site-packages\browser_cookie3\__init__.py", line 374, in __get_connection_legacy
    shutil.copyfile(self.__database_file, self.__temp_cookie_file)
  File "C:\Users\abc40\AppData\Local\Programs\Python\Python311\Lib\shutil.py", line 256, in copyfile
    with open(src, 'rb') as fsrc:
         ^^^^^^^^^^^^^^^
PermissionError: [Errno 13] Permission denied: 'C:\\Users\\abc40\\AppData\\Local\\Google\\Chrome\\User Data\\Default\\Network\\Cookies'

Process finished with exit code 1

[nguyenduclong1147]

Try running the terminal with admin and remember to restart Chrome

[shaggy2626]

After restarting Chrome, the implementation is functioning as expected. Currently, requests to Google API are made sequentially, awaiting each response before proceeding to the next. Given this setup, I'm interested to know if you has encountered rate limiting or blocks from Google. The plan is to make 100-200 sequential calls.

There's a segment of code dedicated to updating cookies after each request:

# Update cookies after every request
cookies = []
for k, v in bard.session.cookies.get_dict().items():
    cookies.append({'name': k, 'value': v})

Is this necessary.

Here is the complete updated code:

Click to expand!

````python import requests from bardapi import Bard, SESSION_HEADERS from PIL import Image from PIL import UnidentifiedImageError import pandas as pd import io import json import browser_cookie3 def _extract_bard_cookie() -> dict: cookie_dict = {} cj = browser_cookie3.chrome(cookie_file='C:\\Users\\\AppData\\Local\\Google\\Chrome\\User Data\\Default\\Network\\Cookies', domain_name=".google.com") for cookie in cj: if cookie.name == "__Secure-1PSID" and cookie.value.endswith("."): cookie_dict["__Secure-1PSID"] = cookie.value if cookie.name == "__Secure-1PSIDTS": cookie_dict["__Secure-1PSIDTS"] = cookie.value if cookie.name == "__Secure-1PSIDCC": cookie_dict["__Secure-1PSIDCC"] = cookie.value return cookie_dict def main(): session = requests.Session() # Get token and cookies from Chrome cookie_dict = _extract_bard_cookie() token = cookie_dict["__Secure-1PSID"] # Set your cookies session.cookies.set("__Secure-1PSID", token) session.cookies.set("__Secure-1PSIDCC", cookie_dict["__Secure-1PSIDCC"]) session.cookies.set("__Secure-1PSIDTS", cookie_dict["__Secure-1PSIDTS"]) # Set the session headers session.headers = SESSION_HEADERS # Send an API request and get a response bard = Bard(token=token, session=session) # Continue with the rest of your code... # Load the text file and read the URLs with open('C:\\Users\\\\Downloads\\Shape.txt', 'r') as f: urls = f.readlines() # Get the total number of URLs total_urls = len(urls) # List to store the output data output_data = [] # Set the batch size batch_size = 5 # Iterate over the URLs for i, url in enumerate(urls): url = url.strip() # Remove leading/trailing whitespace (including newline) # Print the current row number print(f"Processing row {i + 1} / {total_urls}") try: # Get the image from the URL response = requests.get(url) response.raise_for_status() # This line will raise an exception if the request failed # Debug: Print the status code and content type of the response print(f"Status code: {response.status_code}") print(f"Content-Type: {response.headers['content-type']}") try: # Open the image and convert it to bytes image = Image.open(io.BytesIO(response.content)) byte_arr = io.BytesIO() image.save(byte_arr, format='JPEG') image_data = byte_arr.getvalue() except UnidentifiedImageError: print(f"Unable to open image from URL: {url}") continue # Ask the Bard API about the image bard_answer = bard.ask_about_image('What is in the image?', image_data) # Add the URL and response to the output data output_data.append({ "url": url, "response": bard_answer['content'] }) except requests.exceptions.RequestException as e: print(f"Error when trying to access {url}: {e}") output_data.append({ "url": url, "response": str(e) }) # Write the output data to a JSON file every batch_size URLs if (i + 1) % batch_size == 0: with open('C:\\Users\\\\Downloads\\ectionalshape.json', 'w') as f: json.dump(output_data, f, indent=4) # Update cookies after every request cookies = [] for k, v in bard.session.cookies.get_dict().items(): cookies.append({'name': k, 'value': v}) # Write any remaining output data after processing all URLs with open('C:\\Users\\\\Downloads\\onalshape.json', 'w') as f: json.dump(output_data, f, indent=4) if __name__ == "__main__": main() ````

[nguyenduclong1147]

I once ran 20 requests at once, nothing happened. You can ask bard, it seems unlimited requests. bard is still in beta. Still, you should be cautious

[shaggy2626]

I once ran 20 requests at once, nothing happened. You can ask bard, it seems unlimited requests. bard is still in beta. Still, you should be cautious

I was running through intelli J one request after the other and at around 80-90 i got blocked, now even when i try directly through the Bard web UI directly it doesnt go through. i tried another google account and it works but when im using the script im getting stopped.

[nguyenduclong1147]

Probably because of my area, you can ask bard or find bard policy

[umaisyaqoob]

SNlM0e value not found in response. Check __Secure-1PSID value.

This error resolve with clear cookies and repost in own code So your code is runing smooth

[hharera]

My Solution to this issue

import os
import requests
from bardapi import Bard
from dotenv import load_dotenv
from bardapi import BardCookies

load_dotenv('../.env')
token = os.getenv('bard_token')
token_ts = os.getenv('bard_token_ts')

session = requests.Session()
session.cookies.set("__Secure-1PSID", token)
session.cookies.set("__Secure-1PSIDTS", token_ts)
session.headers = {
  "Host": "gemini.google.com",
  "X-Same-Domain": "1",
  "User-Agent": "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.114 Safari/537.36",
  "Content-Type": "application/x-www-form-urlencoded;charset=UTF-8",
  "Origin": "https://gemini.google.com",
  "Referer": "https://gemini.google.com/app",
}
bard = Bard(token=token, session=session)
answer = bard.get_answer("나와 내 동년배들이 좋아하는 뉴진스에 대해서 알려줘")['content']
print(answer)