A 133 KB image pasted into my prompt as text cost 122,584 tokens A developer measured how base64-encoded files embedded in prompts inflate token counts, finding that PNG images and random bytes both consume roughly 914–931 tokens per kilobyte of original file size. A 133 KB social preview image pasted as text cost 122,584 tokens, and the developer estimated a support bot processing 1,000 emails a day with a 40 KB signature logo would see monthly input costs rise from $2.40 to $225.95 on gpt-5.6-mini. The recommended fix is stripping base64 blobs from text before it reaches the model and replacing them with a short note. Here's a mistake that doesn't look like one. You build a support bot that reads incoming emails and drafts replies. The emails are HTML, so you send the HTML. Most of them have the company logo in the signature, embedded right in the markup. Nothing breaks, the drafts are fine, and the input bill is about a hundred times what you estimated. The logo is the problem. When a file travels inside text, it's usually written as base64, and the model has to read every character of it. A token is a chunk of text the model reads; APIs bill per million of them. A tokenizer is the program that cuts text into tokens; I used tiktoken with o200k base , the one OpenAI's current models use. Input tokens are everything you send in the request. Base64 is a way to write any file, an image, a PDF, a zip, using only 64 safe characters letters, digits, + and / , so it can sit inside text formats. A data URI is a link that contains the file itself, like src="data:image/png;base64,iVBORw0KGgo..." in an HTML