Notifications
Clear all
Topic starter
Hi Makers,
I'm trying to update the GPT Assistants in my callin.io workflows and encountering a [400] error. It states that 'gpt-4o' cannot be used with the Assistants API in v1.
On the GPT assistants playground interface, the API is up-to-date and should be callable within the workflow.
Here's the error I'm seeing:
The operation failed with an error. [400] The requested model ‘gpt-4o’ cannot be used with the Assistants API in v1. Follow the migration guide to upgrade to v2: https://platform.openai.com/docs/assistants/migration
Has anyone found a solution for this?
Cheers
Posted : 15/05/2024 11:38 pm
Welcome to the callin.io community!
GPT-4o is still not available for use with Assistants in callin.io yet.
– request private consultation
Join the unofficial callin.io Discord server to chat with us!
Posted : 15/05/2024 11:51 pm
Topic starter
Thanks for the quick heads up.
Posted : 15/05/2024 11:56 pm
It appears the HTTP request needs to be adjusted to send data to the V2 endpoint at https://api.openai.com/v2/assistant .
Posted : 17/05/2024 2:22 am
That's a fantastic suggestion! I'm confident many users would appreciate having this functionality.
You can submit this idea on the Idea Exchange, under App improvements.
Remember to search first to ensure it hasn't already been suggested, preventing duplicate entries.
Don't forget to upvote the feature request to help callin.io support prioritize it!
– request private consultation
Join the unofficial callin.io Discord server to chat with us!
Posted : 17/05/2024 4:07 am
Does this imply that once the callin.io module is updated, we'll need to manually adjust all the “Message an Assistant” modules across our callin.io automations?
Furthermore, it appears this update process is taking a considerable amount of time. Should we anticipate waiting a week (or possibly longer) each time OpenAI updates its model before we can utilize the new versions in our callin.io automations?
Posted : 20/05/2024 8:50 am
OpenAI did not just update their model, but altered the method by which external servers connect and initiate requests to it through the API.
Consequently, it requires time to implement and validate the recent modifications to the modules.
Posted : 20/05/2024 9:06 am
This needs adding as soon as possible. I understand that development can take time, but the current delay is excessive, especially given the demand. Users don't want to implement workarounds to achieve this functionality. Please expedite this. callin.io
Posted : 20/05/2024 4:40 pm
Thanks. I'll monitor this thread.
Posted : 20/05/2024 5:12 pm
Thanks for the heads up. Looking forward to an update soon.
Posted : 20/05/2024 11:04 pm
Hello everyone
I just wanted to confirm that this is indeed a known issue with the "4o" model. Our development team is aware of this and is actively working on updating our ChatGPT module to the most recent API version.
We will notify you once the update has been deployed.
Thank you very much for your patience and understanding.
As a temporary workaround, you could utilize an older assistant model or employ the “Make an API Call” module with a custom API call to the OpenAI API, including the “OpenAI-Beta: assistants=v2” header.
The guide below may offer general assistance on using the “Make an API Call” modules.
How to make an API call: tutorial
Posted : 21/05/2024 5:51 am
Hi everyone, we've updated the Assistant Module within callin.io to support both GPT-4o and functions. You can find more details in this post: Open AI Assistant Module with GPT-4o and functions
Posted : 22/05/2024 7:25 pm
Hi, it's been 4 days already. How far along is this?
Posted : 25/05/2024 1:49 pm
Yeah, burning module executions while polling for my run status isn't a great solution.
Posted : 25/05/2024 2:43 pm
Any ETA on when this functionality will be available?
Could you please provide an estimated timeline for when this will be operational?
Posted : 27/05/2024 3:54 am
Page 1 / 2
Next