AI Rate Card

Claude Sonnet 3.7 vs Claude Haiku 4.5 (latest): API price and specs

Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on input price.

Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on output price.

Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on a 1M-in / 200k-out workload.

FieldClaude Sonnet 3.7Claude Haiku 4.5 (latest)
Lowest input, per 1M$2.60 +420%$0.50
Median input, per 1M$3.00 +200%$1.00
Lowest output, per 1M$13.00 +420%$2.50
Median output, per 1M$15.00 +200%$5.00
Lowest unit price—$0.001 / request
Cost of 1M in + 200k out$5.20 +420%$1.00
Context200K200K
Max output128K64K +50%
Providers5 +85.3%34
RegionsUSAPAC, AU, EU, GLOBAL, IN, JP, US
Acceptsimage, pdf, textapplication/pdf, file, image, image/gif, image/jpeg, image/png, image/webp, pdf, text
Returnstexttext
TypeText modelsText models
Free optionnono
Cheapest providerpoeopenrouter

Your own workload: how many tokens the model gets (input) and gives back (output) - for example, in a month. The "Cost" row of the table above is recalculated for these numbers, for each model.

Computed from current offers each time the page opens; the best value of a row is bold, the percent is how far behind the best.

Providers selling both

deepinfra, edenai, helicone, poe, zenmux

Frequently asked questions

Which is cheaper, Claude Sonnet 3.7 or Claude Haiku 4.5 (latest)?

Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on input price. Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on output price. Claude Haiku 4.5 (latest) is 81% cheaper than Claude Sonnet 3.7 on a 1M-in / 200k-out workload.

Which has the larger context, Claude Sonnet 3.7 or Claude Haiku 4.5 (latest)?

Claude Sonnet 3.7: 200K tokens; Claude Haiku 4.5 (latest): 200K tokens.

Where can I use both Claude Sonnet 3.7 and Claude Haiku 4.5 (latest)?

deepinfra, edenai, helicone, poe, zenmux