Run many prompts through one loaded Aspose.LLM model — amortize the load Cost, pick session-per-prompt vs shared-session patterns....without paying the model-load cost for each one. This is the right... summarization, or any bulk task over a fixed set of inputs....