I hope this message finds you well. I am currently using the most affordable model for neural generation. Additionally, I purchased an extra 1,700 tokens to ensure sufficient coverage for my tasks.
However, I have encountered a significant issue during the design generation process. For what I would consider minor edits — specifically, changing the title in the website header and adding two blocks — the system deducted 1,000 tokens. This seems disproportionately high, especially since these changes introduced no substantial new content or complexity.
Furthermore, I am experiencing persistent loading issues during the process, which further complicates the workflow.
I kindly request that you review this matter and clarify the token consumption logic for such minor adjustments. If this deduction was indeed an error, I would appreciate a correction and possibly a refund of the overcharged tokens.
Please let me know if you need any additional details, such as project IDs or timestamps, to assist with your investigation.
I look forward to your response.