Critical Issues with New 4.1 Models Causing Costly Mistakes

the new 4.1 models do not work, They default to gpt4o which is WAY more expensive We just wasted so much money because the wrong model ID is getting passed in to the OpenAI column Literally could have been a $10k mistake if you guys had accidentally used something like o1-pro or some other model Kinda crazy that there's egregious frontend issues being replicated at scale when your customers have their API keys hooked up to consumption based providers

3 comments

Channeled
APP
·
·
We apologize for the confusion with the model selection. By default, our AI integration uses the Clay-managed account, which can involve credit charges ranging from 0.5 to 6 credits depending on the model. To prevent unexpected costs, you can select your own API key for specific models. Here's how: 1. When setting up the AI integration, click on "Manage Accounts" next to the account field. 2. Follow the prompts to input your OpenAI API key. 3. For Claygent (AI Web Researcher), you can explicitly choose your model in the "Model" field. If you need further assistance with this issue, please request human support by clicking on "Talk to a person" in this chat.
Sources: Claygent (AI Web Researcher) with Artificial Intelligence Use AI integration overview | Documentation | Clay University
Bruno R.
·
·
Hi Chris, thanks for your message - very sorry for this investigating this now. Can you share the url to the table where the defaulting to 4o was happening? Will make sure to take a closer look at what may be happening here.
Channeled
APP
·
·
This thread was picked up by our in-app web widget and will no longer sync to Slack. If you are the original poster, you can continue this conversation by logging into https://app.clay.com and clicking "Support" in the sidebar. If you're not the original poster and require help from support, please post in 02 Support.

Critical Issues with New 4.1 Models Causing Costly Mistakes | Clay