A field guide to AI pricing
Every AI API bills you per million tokens. Nobody tells you how much that actually is. Here's the answer — in novels, photos, podcasts, and dollars.
The scale — input tokens
Input tokens are what you send to the model — documents, code, images, conversations. Here's what a million of them looks like.

1 block = 750 words
≈ 750,000 words
English words
On average one token is about ¾ of an English word, so a million tokens works out to roughly three quarters of a million words.
basis: 1 token ≈ 0.75 words

1 block = 1 page
≈ 1,500 pages
single-spaced pages
At about 500 words per single-spaced page, 750,000 words fills roughly 1,500 pages — a stack of paper about 15 cm (6 in) tall.
basis: 500 words per page

1 block = 4,000 characters
≈ 4,000,000 characters
characters of text
One token is roughly 4 characters of English text, so a million tokens is about 4 million characters — enough to fill 8,000 standard text messages.
basis: 1 token ≈ 4 characters

1 block = 10 paragraphs
≈ 10,000 paragraphs
paragraphs of text
An average paragraph is about 75 words (~100 tokens). A million tokens is roughly 10,000 paragraphs — a library of essays.
basis: ~75 words per paragraph

1 block = 50 sentences
≈ 50,000 sentences
sentences
The average English sentence is about 15 words (~20 tokens). A million tokens is roughly 50,000 sentences.
basis: ~15 words per sentence

1 spine = 1 novel
≈ 8 novels
average novels
A typical novel runs about 90,000 words (~120,000 tokens), so a million tokens holds roughly eight of them.
basis: 90,000 words per novel

the 7 books — filled part = 1M tokens
≈ 69% of Harry Potter
of the complete 7-book series
The full Harry Potter series is 1,084,170 words — about 1.45 million tokens. A million tokens gets you past Order of the Phoenix and into the early chapters of Half-Blood Prince.
basis: Series total 1,084,170 words

2 copies — filled part = 1M tokens
≈ 1.3 × War and Peace
copies of War and Peace
Tolstoy’s epic is about 587,000 words in English translation (~780,000 tokens). A million tokens covers it with room for a second read of the first third.
basis: 587,000 words per copy

2 copies — filled part = 1M tokens
≈ 1.6 × Lord of the Rings
copies of the trilogy
Tolkien's trilogy is about 481,000 words (~641,000 tokens). A million tokens covers the full trilogy with room for more than half a reread.
basis: 481,000 words (trilogy without appendices)

1 Bible — filled part = 1M tokens
≈ 96% of the Bible
of the King James Bible
The King James Bible is about 783,000 words (~1.04 million tokens). A million tokens gets you through almost the entire text — falling just short at Revelation.
basis: 783,137 words (KJV)

the 5 books — filled part = 1M tokens
≈ 42% of Game of Thrones
of the 5 published ASOIAF books
George R.R. Martin's A Song of Ice and Fire is about 1,770,000 words (~2.36 million tokens). A million tokens covers the first two books and about 30% of A Storm of Swords.
basis: 1,770,000 words (5 published books)

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

1 tile = 1 image
≈ 1,000 images
detailed images analyzed
Vision models spend roughly 800–1,600 tokens to read one detailed image, so a million tokens lets a model look at about a thousand photos or screenshots.
basis: ~1,000 tokens per image

1 block = 1 PDF page
≈ 650 PDF pages
visually-processed PDF pages
When a model reads a PDF page as both text and image (preserving layout, tables and figures), each page costs ~1,500 tokens. Text-only extraction is cheaper: ~2,000 pages per million tokens.
basis: ~1,500 tokens per page with layout

1 block = 25 messages
≈ 75,000 chat messages
chat messages
A typical text message is about 10 words (~13 tokens). A million tokens is over six years of texting at 30 messages a day.
basis: ~13 tokens per message

1 bar = 1 hour of talking
≈ 83 hours of speech
hours of transcribed speech
People speak about 150 words per minute (~200 tokens/min), so a million tokens transcribes about 83 hours — three and a half days of non-stop talking.
basis: 150 words per minute

1 block = 1 episode
≈ 110 podcast episodes
45-minute episodes
At ~9,000 tokens per 45-minute episode transcript, a million tokens holds about 110 episodes — a full year of a weekly show, twice over.
basis: 45 min per episode, 150 wpm

1 block = 10 posts
≈ 14,000 max-length posts
280-character posts
A maxed-out 280-character post is about 70 tokens, so a million tokens is roughly 14,000 of them — about 10 years of posting four times a day.
basis: 280 chars ≈ 70 tokens

1 block = 10 emails
≈ 10,000 work emails
typical work emails
The average work email body runs about 75 words (~100 tokens). A million tokens is ten thousand of them — several years of a busy inbox.
basis: ~100 tokens per email

1 spine = 1 screenplay
≈ 37 movie screenplays
feature-film screenplays
A feature screenplay is around 20,000 words (~27,000 tokens), so a million tokens holds about 37 movies’ worth of scripts.
basis: 20,000 words per screenplay

1 block = 1 T&C agreement
≈ 125 Terms & Conditions
Terms & Conditions agreements
The average social media Terms & Conditions is about 6,000 words (~8,000 tokens). A million tokens is 125 of them — still less than what you've blindly agreed to this year.
basis: ~6,000 words per T&C

1 block = 10 manuals
≈ 15,000 IKEA manuals
IKEA instruction manuals
An IKEA manual averages about 50 words of actual text (~67 tokens) — they're mostly pictures of confused stick figures. A million tokens is 15,000 of them. You'd still build the shelf wrong.
basis: ~50 words per manual

1 block = 10 bios
≈ 50,000 Tinder bios
dating app bios
The average dating profile bio is about 15 words (~20 tokens) — mostly "love to travel" and dog photos. A million tokens is 50,000 of them. That's a lot of swiping.
basis: ~15 words per bio

1 block = 1 article
≈ 1,000 Wikipedia articles
average Wikipedia articles
The average Wikipedia article is about 725 words (~970 tokens), so a million tokens is roughly a thousand articles of general knowledge.
basis: ~725 words per article

1 bar = 2 songs
≈ 2,500 songs
songs’ worth of lyrics
Song lyrics average ~300 words (~400 tokens), so a million tokens is the lyrics to about 2,500 songs — over six days of continuous music.
basis: ~300 words per song
The other side — output tokens
Output tokens are what the model generates back. They cost 2-8x more than input tokens, but each response is typically short — a few hundred to a few thousand tokens. A million output tokens accumulates over many requests.

1 block = 1 response
≈ 2,000 detailed responses
long AI responses
A detailed AI response (explanation, analysis, essay) averages about 500 tokens. A million output tokens is roughly 2,000 of these — months of heavy daily usage.
basis: ~500 tokens per detailed response

1 block = 1 completion
≈ 5,000 code completions
function implementations
A typical code completion — a function implementation, a refactor, a bug fix — runs about 200 tokens. A million output tokens is 5,000 of these.
basis: ~200 tokens per code completion

1 block = 1 article
≈ 500 blog posts
AI-written articles
A full blog post or report generated by AI averages about 2,000 tokens (~1,500 words). A million output tokens produces 500 complete articles.
basis: ~2,000 tokens per article

1 block = 10 emails
≈ 10,000 email drafts
AI-drafted email replies
A professional email reply generated by AI averages about 100 tokens (~75 words). A million output tokens drafts 10,000 emails — years of inbox assistance.
basis: ~100 tokens per email draft

1 block = 1 summary
≈ 1,000 document summaries
page-long summaries
A thorough document summary runs about 1,000 tokens (~750 words, a full page). A million output tokens produces 1,000 of these — an entire research library summarized.
basis: ~1,000 tokens per summary

1 block = 10 replies
≈ 50,000 chatbot replies
short chatbot responses
A quick chatbot answer — a classification, a yes/no, a short extraction — averages about 20 tokens. A million output tokens is 50,000 of these, enough for a busy customer service bot.
basis: ~20 tokens per short reply
Why output costs more
Input tokens are cheap because the model just reads them. Output tokens are expensive because the model has to generate each one sequentially — predicting the next word thousands of times. That's why most models charge 3-6x more for output.
The price
Pick a budget and see how many novels' worth of text each model writes (or reads) for it. Prices updated 2026-07-28.
One spine = one novel written for $10. Same money, 60× difference between the cheapest and priciest model. Filtering the scale section above switches the unit here too.
| Model | Provider | Input / 1M | Output / 1M | Context |
|---|---|---|---|---|
| GPT-5.6 Sol | OpenAI | $5 | $30 | 1.05M |
| GPT-5.5 Pro | OpenAI | $30 | $180 | 400K |
| Claude Fable 5 | Anthropic | $10 | $50 | 1M |
| GPT-5.6 Terra | OpenAI | $2.50 | $15 | 1.05M |
| GPT-5.5 | OpenAI | $5 | $30 | 400K |
| Claude Opus 4.8 | Anthropic | $5 | $25 | 1M |
| Claude Sonnet 5 | Anthropic | $2 | $10 | 1M |
| Gemini 3.1 Pro | $2 | $12 | 1M | |
| Gemini 3.5 Flash | $1.50 | $9 | 1M | |
| GPT-5.6 Luna | OpenAI | $1 | $6 | 1.05M |
| Claude Haiku 4.5 | Anthropic | $1 | $5 | 200K |
| Gemini 3 Flash | $0.50 | $3 | 1M |
GPT-5.5: Long-context (>272K input) priced at 2× input / 1.5× output. Claude Sonnet 5: Introductory pricing — list price is $3 / $15 from Sep 1, 2026. Gemini 3.1 Pro: Prices double for prompts above 200K tokens ($4 / $18).
The calculator
Pick a model, set a budget, and see what it translates to.
1,000,000tokens
The fine print
A token is the unit AI models read and write text in — usually a word, part of a word, or a punctuation mark. "Cat" is one token, but "tokenizer" splits into pieces like "token" + "izer". In typical English, one token is about ¾ of a word, so 1,000 tokens ≈ 750 words.
Generating text is much more computationally expensive than reading it. Every provider reflects that: output tokens typically cost 5–6× more than input tokens. Your real bill depends on your mix — a summarization app sends lots of input and gets little output, while a writing app is the opposite.
No. Each model family uses its own tokenizer, so the same text can differ by 30% or more in token count between providers — and even between model generations from the same provider. That means a per-million-token price is only comparable once you account for how each tokenizer counts your text.
Every major provider offers a token-counting endpoint or tokenizer tool (Anthropic has a count-tokens API, OpenAI and Google publish tokenizers). Rough planning is easier: divide your word count by 0.75, or your character count by 4.
Anywhere from $0.50 (Gemini 3 Flash input) to $180 (GPT-5.5 Pro output) as of July 2026 — a 360× spread. That is exactly why it pays to know what a million tokens is before choosing a model.
Our methodology
Token conversion
All comparisons use the widely-accepted standard of 1 token ≈ 0.75 English words (approximately 4 characters). Every figure on this site shows its underlying assumption so you can verify the math yourself.
Pricing data
Prices are verified against each provider's official API documentation and updated regularly. The date of the last pricing check is displayed alongside every price on this site.
Reference sources
Word counts for specific works (novels, screenplays, religious texts) are sourced from Wikipedia, Project Gutenberg, and other authoritative public references. Each comparison links to its source.
Independence
This site is independently maintained and not affiliated with any AI provider — not OpenAI, not Anthropic, not Google. We have no financial relationship with any model provider.
We track prices daily. Get a one-time email when your model gets cheaper.