Troubleshooting AI Errors Related to Credit Budgets

When using AI, I'm having an issue recently where a considerable percentage of prompt responses are mentioning errors or not being able to perform due hitting credit budgets. If I change the model or increase the budget the same error happens just indicating a higher credit spend. I have ran identical prompts (for analysis and respective messaging generation) with the same model (Claude 4.5 Sonnet) in the past and have not experienced this huge amount of errors. Is there a way to avoid this issue? As example, in a sample table of 80 rows, I've got 31 with errors. Thanks