Chunking/token limit bug still present on long posts

Hi Elliot,

I’m hitting the same chunking/token limit bug that was reported in the support thread from December 2025 (“Error with generating”). The error message in that thread was “Input of 2209 tokens is over the maximum input limit of 2000 tokens” but on my end the plugin just shows a generic “An error occurred while generating the TTS” alert with no details.

I’m on the latest version of the plugin using the OpenAI API. The issue only affects longer posts (mine are around 60,000 characters). Shorter posts around 15,000-20,000 characters generate fine.

I did a binary search to narrow it down. The text itself is clean (no special characters, no encoding issues). Any individual section works on its own. But when the total content hits a certain length, the plugin’s internal chunking appears to produce at least one chunk that exceeds the 2,000 token limit, and the whole generation fails.

For now I’m manually splitting my content into chunks under 7,000 characters and generating them separately, but obviously that’s not a sustainable workflow.

You mentioned in that December thread that a fix was coming in the next update. Is there a timeline on that? Happy to help test if you have a beta version.

Thanks,

The page I need help with: [log in to see the link]