Paste your prompt or document to estimate its token count and see the cost across the major
models — side by side. Everything runs in your browser.
01 Your text
Characters: 0Words: 0
assume output = %
of input (for blended cost)
02 Estimate
0tokens
~ estimated from text length · not an exact tokenizer count
Model
Rate /1M (in/out)
Est. cost
How this works. Token counts are estimated from character and word counts using per-family
heuristics (roughly 3.5–4 characters per token for English) — not each provider's exact
tokenizer, which would require a backend. Good enough for estimating cost and context fit; for exact counts,
use the provider's own tokenizer. Cost = (input tokens × input rate) + (output tokens × output
rate). Prices are standard pay-as-you-go rates per 1M tokens, gathered June 2026 — always confirm
against the provider's pricing page before relying on them. Caching and batch discounts not included.