What will this prompt cost?

Pick the model you use, paste a prompt, and see the bill — then what else you could have sent it to for the same money. Nothing leaves this tab.

your model
65 input tokens likely 55–78 222 characters 38 words 3.4 chars a token

Expecting 900 output tokens — the prompt asks for something itemised, which scales with what it finds. It is a guess from how the prompt is written; type over it if you know better.

Every published rate

Pick your model above to see it against its neighbours. Until then, the whole field, cheapest first.

modelper call× 1overall
Ministral 3 14B Reasoning Mistral AI$0.00019$0.0001947
DeepSeek V4-Flash DeepSeek$0.00026$0.0002658
MiMo-V2.6-Flash Xiaomi$0.00026$0.00026—
Devstral Small 2 Mistral AI$0.00028$0.0002844
Qwen3.8-Flash-Next Alibaba$0.00043$0.0004361
Claude Haiku 5.5 Anthropic$0.00046$0.0004665
GPT-6 Luna OpenAI$0.00046$0.00046—
GLM-5.3-Flash Z.ai$0.00046$0.00046—
Qwen3 32B Alibaba$0.00059$0.0005934
DeepSeek V4-Pro DeepSeek$0.00081$0.0008166
MiMo-V2.6-Pro Xiaomi$0.00081$0.00081—
DeepSeek-V4.1-Flash DeepSeek$0.00110$0.0011059
MiniMax M2 MiniMax$0.00110$0.0011035
MiniMax M2.1 MiniMax$0.00110$0.0011049
MiniMax M3 MiniMax$0.00110$0.0011062
Gemini 3.1 Flash-Lite Google$0.00137$0.0013737
Mistral Large 3 Mistral AI$0.00138$0.0013840
Devstral 2 Mistral AI$0.00183$0.0018348
GLM-4.5 Z.ai$0.00202$0.0020247
GLM-4.6 Z.ai$0.00202$0.0020234
GLM-4.7 Z.ai$0.00202$0.0020253
Gemini 2.5 Flash Google$0.00227$0.0022730
Gemini 3.5 Flash-Lite Google$0.00227$0.0022748
Qwen3 235B A22B Alibaba$0.00257$0.0025741
Gemini 3 Flash Google$0.00273$0.0027353
Qwen3.8-27B Alibaba$0.00273$0.0027356
Kimi K2.5 Moonshot AI$0.00274$0.0027458
GLM-5 Z.ai$0.00295$0.0029551
Qwen3.5 397B A17B Alibaba$0.00328$0.0032863
DeepSeek-V4-Pro-0813 DeepSeek$0.00365$0.0036557
Kimi K2.6 Moonshot AI$0.00366$0.0036662
Kimi K2.7-Code Moonshot AI$0.00366$0.00366—
Mistral Large 4 Mistral AI$0.00385$0.00385—
Muse Spark 1.2 Meta$0.00391$0.00391—
Muse Spark 1.3 Meta$0.00391$0.00391—
o4-mini OpenAI$0.00403$0.00403—
GLM-5.1 Z.ai$0.00405$0.0040550
GLM-5.2 Z.ai$0.00405$0.0040564
GLM-5.3 Z.ai$0.00405$0.00405—
Claude Haiku 4.5 Anthropic$0.00456$0.0045643
Magistral Medium Mistral AI$0.00463$0.0046333
Grok 4.5 xAI$0.00553$0.00553—
Grok 4.6 xAI$0.00553$0.00553—
Grok 4.7 xAI$0.00553$0.00553—
Mixtral 8x22B Mistral AI$0.00553$0.00553—
Qwen3.8-Max (Qwen3.8-2.4T-A95B) Alibaba$0.00553$0.0055369
Gemini 3.6 Flash Google$0.00685$0.0068552
Mistral Medium 3.5 Mistral AI$0.00685$0.0068557
GPT-4.1 OpenAI$0.00733$0.0073333
o3 OpenAI$0.00733$0.00733—
Gemini 3.5 Flash Google$0.00820$0.0082052
Gemini 2.5 Pro Google$0.00908$0.0090843
GPT-5 OpenAI$0.00908$0.0090859
GPT-5.1 OpenAI$0.00908$0.0090862
GPT-5.1-Codex-Max OpenAI$0.00908$0.0090858
Claude Sonnet 5.5 Anthropic$0.00913$0.0091373
GPT-6.1 Sol OpenAI$0.00913$0.00913—
GPT-6 Sol OpenAI$0.00913$0.00913—
GPT-4o OpenAI$0.00916$0.00916—
Gemini 3.1 Pro Google$0.0109$0.010969
GPT-5.2 OpenAI$0.0127$0.012774
GPT-5.2-Codex OpenAI$0.0127$0.0127—
GPT-5.3-Codex OpenAI$0.0127$0.0127—
GPT-5.4 OpenAI$0.0137$0.013763
Claude Sonnet 4 Anthropic$0.0137$0.013738
Claude Sonnet 4.5 Anthropic$0.0137$0.013756
Claude Sonnet 4.6 Anthropic$0.0137$0.013762
Claude Sonnet 5 Anthropic$0.0137$0.013769
Claude Opus 5.5 Anthropic$0.0183$0.018375
Claude Opus 4.5 Anthropic$0.0228$0.022860
Claude Opus 4.6 Anthropic$0.0228$0.022867
Claude Opus 4.7 Anthropic$0.0228$0.022873
Claude Opus 4.8 Anthropic$0.0228$0.022876
Claude Opus 5 Anthropic$0.0228$0.022878
GPT-5.5 OpenAI$0.0273$0.027363
GPT-5.6 OpenAI$0.0273$0.027373
Claude Fable 5 Anthropic$0.0456$0.045672
Claude Fable 5.1 Anthropic$0.0456$0.045673
Claude Mythos 5 Anthropic$0.0456$0.045681
GPT-6 Astra OpenAI$0.0456$0.045674
o1 OpenAI$0.0550$0.055035
Claude Opus 4 Anthropic$0.0685$0.068541

how this counts

  • It is an estimate, and deliberately a range. An exact count needs the tokenizer the model was trained with — a different vocabulary of a hundred thousand merges per family, megabytes each. Shipping one and quoting it for the whole table would look precise while being wrong for most of it. The same token count is used for every model here for the same reason.
  • What it does instead is count structure. Short words are one token; long ones split roughly every four characters; digits break into far smaller pieces than letters; punctuation is usually its own token; and CJK, which has no spaces, runs closer to one token per character.
  • The output box is the one that matters. Output costs several times input almost everywhere, and a reasoning model charges for thinking tokens you never read. A long answer from a cheap model can cost more than a short one from an expensive model.
  • A cheaper neighbour is not automatically a better buy. The overall column is there so you can see what you would be giving up, and a model with no ranking never appears as a recommendation — only as a neighbour.
  • Prices are the vendor's standard pay-as-you-go rate as of 2026-07-25, excluding batch, cached-input, off-peak and priority tiers. Open weights with no first-party API are not priced here: what they cost is your own hardware.