Gemini models truncate chat history at 32,000 characters?

[nk121]

I'm using VertexAI API to use Gemini 1.5 models. The documentation on this page:

https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/function-calling

At the bottom, it has this line:

"The history of the conversation includes the queries, the function calls, and the function responses of previous conversation turns. Vertex AI truncates the history of the conversation at 32,000 characters."

Is that actually true for the newer Gemini 1.5 models with the 1M context limit? And is that really characters and not tokens?