Honest analysis of LLM pricing, tokenization quirks, and the actual cost of running models in production.
OpenAI shipped GPT-5.5 with a quiet 2× price hike on input and output. Both models use the same tokenizer. Here's what that does to a 30,000-request-per-month workload, and when it's worth paying.