Need this sooo bad.

GPT actions in workflows have such a limited range of models. Doesn't make sense. Why are the GPT 4.1 and other models not available? GPT 4.1 perform really well and have a 600k larger context window than GPT 5 Models and an 872k larger context window than GPT 4o. We have just done 1000's of split test across different models and found the GPT 4.1 mini to perform really well and super cost effective.

The context window becomes super important when we are chaining a series of outreach messages together and brining in context / memory.