A million tokens, measured in
Code is denser than prose per line but shorter per idea: a million tokens is roughly 100,000 lines — a real mid-sized codebase, or the original Doom engine twice over.
1 block = 100 lines of code
≈ 100,000 lines of code
lines of source code
Code averages roughly 10 tokens per line, so a million tokens is about 100,000 lines — nearly twice the original Doom engine (~57,000 lines).
basis: ~10 tokens per line
Code tokenizes differently from prose. Whitespace, brackets, and camelCase identifiers all affect the token count, and most code averages about 10 tokens per line compared to prose at roughly 20 tokens per sentence. For engineering teams evaluating AI code assistants, the question is straightforward: can the model see enough of your codebase in a single pass to give useful answers? A hundred thousand lines covers a mid-sized production service.
Everything on this page is exactly one million tokens. So the price of having a model read — or write — all of it is simply each model's per-million rate. Updated 2026-07-28.
| Model | Read all of it (input) | Write all of it (output) |
|---|---|---|
| GPT-5.6 Sol | $5 | $30 |
| GPT-5.5 Pro | $30 | $180 |
| Claude Fable 5 | $10 | $50 |
| GPT-5.6 Terra | $2.50 | $15 |
| GPT-5.5 | $5 | $30 |
| Claude Opus 4.8 | $5 | $25 |
| Claude Sonnet 5 | $2 | $10 |
| Gemini 3.1 Pro | $2 | $12 |
| Gemini 3.5 Flash | $1.50 | $9 |
| GPT-5.6 Luna | $1 | $6 |
| Claude Haiku 4.5 | $1 | $5 |
| Gemini 3 Flash | $0.50 | $3 |
From $0.50 to $180 for the same million tokens. Try your own budget in the calculator →
Or start over with the full picture — scale, prices, and the budget calculator.