Increase timeout for API requests

It looks like you implemented a new timeout function which breaks my API calls.

Where can I increase the timeout?

same thing happening for me. Also, seems like it is faster than 10 second for me. Breaking every typebot i have with AI calls

Example: "response": {
"statusCode": 500,
"data": {
"message": "Error from Typebot server: TimeoutError: Timeout awaiting 'response' for 10000ms"
}
}

Happens when I pass anything over 500 words to the AI

happens more when I make calls to Anthropic to https://api.anthropic.com/v1/complete {
"prompt": "\n\nHuman: {{ProposalGPT}}. Here is the information to build the proposal from: {{User Messages}}, Timeline: {{timeline}} . Today's date is {{todayis}}. This is your chat memory:{{Message History}} \n\nAssistant:",
"model": "claude-2.1",
"max_tokens": 100000,
"max_tokens_to_sample": 100000,
"temperature": 0.5,
"top_p": 1.0,
"frequency_penalty": 0.5,
"presence_penalty": 0.0,
"stop_sequences": ["Human:", "Assistant:"]
}

Testing the API call, I get status 200. So this could be a problem with the way the data is being pushed to a variable. Note this has worked for months until the other week

This is still an issue for anthropic api calls

You mean the Error from Typebot server: TimeoutError: Timeout awaiting 'response' for 10000ms still happening for the Anthropic call?

That should not happen. Do you have any other webhook that could trigger this message?

Will it take a long time to increase? Is there any palliative method?

@Baptiste

Putting the domain of my external services, will it work? I'll test it today and get back here

Check the possibility of being configured through environment variables. I use caprover and I think it would be wonderful.

I receive timmout from my service Flowise.

Hey @Igor

You should not have a timeout when self-hosting by default (as long as the CHAT_API_TIMEOUT env is not defined). What version are you running?

@Ilja Did you manage to disable the timeout on your self-hosted instance?

First of all thank you for the fast repsonse and fixing for the timeout 2 days ago. We have tried the values I sent you (empty, 120000ms, 0 and false false. Couldn't get it work with these values. As it was critical and we got a proxy solution later. We will try now to remove the values - hopefully it will work.

we tried it: unfortunately it didnt work

Ok will try to reproduce the issue

Are you sure you have deployed version 2.21.2?

Can you describe the error? How do you test it?

the env variable was removed:

but still the timeout 10s issue persisted:

instead of checking if timeout variable is defined or not: wouldn't it be better to define the default timeout value using the variable?

Ooooh right, it wasn't enabled for test requests.

I'll push a new version for this ASAP

Released 2.21.3 👍

@Baptiste Through Postman, I use an OpenAI conversation completion block and create audio, but through the API in a Postman call there is a limit of just 60 seconds?

I use new version 2.21.3

This is not a timeout issue, it just means your server can't be reached