I counted tokens for the same data in 7 formats. Pretty JSON costs 3 CSV

A developer measured token counts for the same 20-row, 5-field product table serialized in seven formats using the o200k_base tokenizer behind GPT-4o, finding pretty-printed JSON costs roughly 3× the tokens of CSV (884 vs 300) and XML 3.63× (1,088 tokens). The same data minified as JSON came to 525 tokens and YAML to 649, while TSV (296) and CSV (300) were cheapest; the older cl100k_base tokenizer produced nearly identical results within 2%. At 30,000 monthly requests with a 20-row table attached, the developer estimated CSV costs about $18/month versus $53 for pretty JSON at $2 per million input tokens, and released a browser-based token counter that runs the same tokenizer locally.

Most of us paste data into LLM prompts as JSON, often straight from JSON.stringify data, null, 2 . I had never checked what that costs in tokens, so I took one table, wrote it in seven formats and counted. Short version: pretty-printed JSON uses about 3× the tokens of CSV. A product table: 20 rows, 5 fields id, name, price, in stock, category . One row in CSV: 1001,Wireless Mouse,9.99,false,electronics Same data, seven formats, counted with o200k base , the tokenizer behind GPT-4o and later OpenAI models: js import { getEncoding } from "js-tiktoken"; const enc = getEncoding "o200k base" ; const count = s = enc.encode s .length; count csv ; // 300 count JSON.stringify rows ; // 525 count JSON.stringify rows, null, 2 ; // 884 | Format | Tokens | vs CSV | |---|---|---| | TSV | 296 | 0.99× | | CSV | 300 | 1.00× | | Markdown table | 373 | 1.24× | | JSON, minified | 525 | 1.75× | | YAML | 649 | 2.16× | | JSON, pretty 2-space | 884 | 2.95× | | XML | 1,088 | 3.63× | The older cl100k base tokenizer gave nearly identical numbers, within 2% for every format. "name": , "price": twenty times. CSV writes them once, in the header. YAML is the odd one: fewer characters than minified JSON, more tokens, because every field gets its own line and its own key. Coding agents send a lot of code, so I tried a few things: | Test | Result | |---|---| | 16-line Python file: 4 spaces vs 2 spaces vs tabs | 135 / 135 / 133 tokens, basically no difference | | Same file without its one-line docstring | 135 → 123 −9% | | 10-line JS function, minified | 92 → 47 −49% | | One UUID | 18 tokens | subtotal and taxRate , which is exactly what helps the model understand the code. Dropping unused fields, null fields and extra decimal places helps too. Say you attach a 20-row table to every request, 1,000 requests a day, 30,000 a month, at $2 per million input tokens: | Format | Tokens / month | Cost / month | |---|---|---| | CSV | 9.0M | $18 | | JSON, minified | 15.8M | $32 | | JSON, pretty | 26.5M | $53 | | XML | 32.6M | $65 | Invisible per request, real for a product, and it scales with bigger tables, RAG results and long API responses. I built a free token counter https://tokensave.app/ for this. It runs the same o200k tokenizer in your browser, so nothing you paste is uploaded. Paste your data in two formats and compare tokens and cost per model. Full write-up with more detail: JSON vs YAML vs CSV: which format uses the fewest tokens? https://tokensave.app/blog/json-vs-yaml-vs-csv-tokens