Services Plugins FAQs

Several Serious Tier1Issues with "Full OpenAI ChatGPT-4" plugin

Help Needed
#1

Several issues with “Full OpenAI ChatGPT-4” plugin. It is delaying our launch and has been painful. Need help fast.

Question 1: Why isn’t there an option to get the response status in real-time in image generation? As it takes forever and is not going to work for anyone.

Question 2: why is the “Full OpenAI ChatGPT-4” plugin giving extremely slow responses? Incredibly slower than the chatGPT playground. It is not OpenAI’s API either, as we tried another plugin with fewer capabilities, but it was similar in speed to the Playground.

Also, the responses are much slower than on your demo using the same prompts.

We need help fast, please!

#2

Here is a screenshot and video of the issue regarding the response status and the slowness of a response.

thumb_Screenshot 2023-11-19 at 12.02.20 AM
.

#3

Here are the duration images that take 1:09 or one minute and 9 seconds versus 7 seconds in your plugin demo.

Start of Prompt
Start of Prompt1079×547 40.1 KB

Finished Image Generation
Finished Image Generation1736×854 97.1 KB

#5

Here are the associated workflows screenshots:

Screenshot 2023-11-19 at 12.24.02 AM
Screenshot 2023-11-19 at 12.24.02 AM2590×640 73.5 KB

Screenshot 2023-11-19 at 12.19.22 AM
Screenshot 2023-11-19 at 12.19.22 AM2646×220 20.6 KB

Image Generation Workflows
Image Generation Workflows1231×788 74.4 KB

#6

Here is the plugin setup:

Full OpenAI ChatGPT -4 Plugin setup
Full OpenAI ChatGPT -4 Plugin setup992×722 78.4 KB

#7

How many days do you have to wait for support on here? We have purchased several plugins and were considering doing the mobile app solution, but given the lack of support probably not.

#8

Hi @mark.lynd,

We apologize for the delay in responding. Typically, we strive to address queries within one working day. However, if the question requires investigation from our development team, it might take longer due to the current workload. Your understanding is greatly appreciated.:pray:

We’re actively investigating these matters with our developer team. I’ll ensure to provide you with the necessary information as soon as possible.

In the meantime, please let me know if you have any further questions. I will be happy to assist you.

Best regards,
Gulce

#9

Hi Gulce,

Thanks for the reply. To be honest, we are seeing several issues with this plugin. On top of the ones already mentioned, we are seeing inconsistent generation times for the same prompt if you use it three times in a row for example it may take 7 seconds for one, 10 seconds for the second run, and then over 1 minute for the next run. This is why a response status is so critical, otherwise user is just waiting and doesn’t have any way to know if it is “hung up’” or not. Not sure why the generation is so inconsistent either, as it is not inconsistent on playground or your demo or the other plugin.

Hopefully, we will here something back form your dev folks as we are holding launch up for this.

Thanks,

Mark

#10

Hi Mark,

I apologize once again for the inconvenience caused. :pray:

Indeed, it’s strange that we’re not encountering the issues you mentioned on our end. To get to the bottom of this and help you out faster, giving us access to your app would be super handy.

If it’s convenient, you can add our support@zeroqode.com email as a collaborator to your app (even if it’s on the Free plan, no worries about any warning messages). Please let me know the page name where the issue is found. You can send the details via DM.

Thanks a bunch!

Best regards,
Gulce

#11

Thank you, I am adding your email now as a collaborator.

#12

You have been added as a collaborator. The page name is: “chat_new”. Thank you.

#13

Were you able to get in ok?

#14

Hi @mark.lynd,

Yes, thank you so much for providing access. I will share it with our developer team and as soon as I have any news, I will get back to you. I appreciate your patience. :pray:

Best regards,
Gulce

#15

Thank you Gulce, as we are delaying launch until we get some resolution on this item.

#16

Hi @mark.lynd

I will do my best to speed up this process. Thank you for understanding. :pray:

Best,
Gulce

#17

Thanks Gulce,

But, seriously, who is in charge there and please escalate, as I am a paying customer and 5 days is absurd for support. I don’t think anyone would buy ZeroQode plugins if they new the product support was so slow and it was posted over on Bubble. You are doing fine, but this process and timing is really poor.

Thanks

#18

Hi @mark.lynd,

I completely empathize with your situation. Our plugin developer is currently juggling multiple tasks with varying levels of priority, which has resulted in some delays. Additionally, investigating your request involves a detailed analysis of response times, use cases, and comparisons, which takes a considerable amount of time. I really appreciate your understanding. :pray:

  • We have set up a test page within the application to evaluate our plugin ‘Full Open AI - Open AI - Chat GPT-3.5-turbo.’ Here are the outcomes we observed when the prompts differ:

First run: 2s 402 ms

image
image1623×493 86.3 KB

Second run:

second run
second run1685×433 68.1 KB

Third run:

3rd
3rd1625×402 68.7 KB

  • The case with the same prompt:

First run:

FIRST2
FIRST21752×365 34.4 KB

Second run:

SECOND2
SECOND21781×394 18.6 KB

Third run:

THIRD2
THIRD21780×414 19.1 KB

  • The plugin to the competitor: ChatGPT Toolkit (with streaming)

Different prompt:

First run: 16 S

third1
third11232×57 21.9 KB

Second run: 12

image (1)
image (1)1230×67 13.1 KB

Same prompt: Error on the last one

image (2)
image (2)1291×135 13.6 KB

  • We’ve also checked the “Open AI - Image generations” action: Steps 7,8,9 need to be optimized.

image gen
image gen1244×207 27.5 KB

image (4)
image (4)1268×424 124 KB

On the server side, new APIs are produced to guarantee security so that they can run slower. So, it’s normal if you notice these responses taking a bit longer.

Regrettably, there isn’t an API available from OpenAI that aligns with what we can implement. We apologize for any inconvenience this may cause.

Feel free to share your thoughts or any concerns you might have. It will be a pleasure to cooperate. Hope to hear from you soon!

Best regards,
Gulce

#19

Thanks Gulce. It is appreciated. Have a question: What does this from your response mean or what action do we need to take, as we are not following: " * We’ve also checked the “Open AI - Image generations” action: Steps 7,8,9 need to be optimized."

#20

Hi @mark.lynd,

Hope you’re doing well!

Sure! Let me clarify it for you. You need to review the action steps for possible speed improvements. Identify any steps causing prolonged processing times and consider adjusting their setup to enhance efficiency. Although it’s not mandatory, optimizing this aspect could significantly speed up the process. Please let me know if you have any additional questions. :pray:

Best regards,
Gulce