How many tokens is a resume?

Resumes tokenize a little heavier than prose because they are dense with names, dates, abbreviations, and formatting — all things BPE tokenizers split aggressively. A 550-word resume that would be ~730 tokens as plain prose often lands closer to 800-900 tokens after bullet structure and jargon.

For AI screening pipelines the math is friendly: even at 1,000 tokens per candidate with a structured 200-token extraction output, screening 10,000 applications costs a few dollars on a budget model. The constraint is quality and fairness of the evaluation, not token spend.

For candidates using AI to tailor a resume, a full tailoring pass (resume in, job description in, revised resume out) runs about 2,500-3,500 total tokens — fractions of a cent on any model tier.