CheckPre

LLM Classification Cost Calculator

What it costs to tag, route or classify large volumes of short text with an LLM, and which small models do it most cheaply.

Updated · By Niraj Kumar · 6 sources

Model

Each request

Count text inCount text in
tokens

System prompt, conversation history, documents and tool definitions are all resent on every call.

tokens

The visible answer.

Count tokens from a prompt or JSON payloadShow

Plain text, OpenAI Chat or Responses JSON, or Anthropic Messages JSON. A final assistant message or a response body counts as output. Counted in your browser; nothing is uploaded.

Loading the tokenizer…

Counts use OpenAI's o200k tokenizer, so they are exact for GPT-6 and GPT-5 models, plus a few tokens of message framing. Other providers' tokenizers count the same text differently and do not publish a conversion factor; Anthropic says Claude models from 4.7 on produce about 30% more tokens than earlier Claude models. For an exact count on another model, paste one of its API responses: the usage it reports is shown above.

tokens

Thinking tokens are billed as output. Leave at 0 for models or settings that do not reason.

Volume

PricePrice

3,000,000 requests over 30 days.

Discounts

%
0%100%

The share of input that repeats exactly at the start of every request, such as a fixed system prompt. The prompt caching calculator prices cache writes and expiry in detail.

Batch APIBatch API
Monthly cost$67.50$0.0000225 a request on GPT-6 Luna
Per 1,000 requests
$0.0225
Input
$60.00
Output
$7.50
Price rank
2nd cheapest of 24
Cheapest listed model
Claude Haiku 5.5
$67.50 a month, $0 less
Output11%
  • Input89%
  • Output11%
Every model on this workload, per month400 tokens in, 10 out, 100,000 a day, batch rates. Hover or focus a row for rates.
  1. 1Claude Haiku 5.5Anthropic$67.50
  2. 2GPT-6 LunaOpenAI$67.50
  3. 3GPT-5.6 LunaOpenAI$138.00
  4. 4Qwen3.8 FlashAlibaba$194.10
  5. 5Gemini 3.5 Flash-LiteGoogle$217.50
  6. 6DeepSeek V4.1 FlashDeepSeek$396.00
  7. 7Gemini 3.8 FlashGoogle$506.25
  8. 8Qwen3.7 PlusAlibaba$528.00
  9. 9Claude Haiku 4.5Anthropic$675.00
  10. 10Grok 4.3xAI$1,260
  11. 11Claude Sonnet 5.5Anthropic$1,350
  12. 12Claude Sonnet 5Anthropic$1,350
  13. 13GPT-6.1 SolOpenAI$1,350
  14. 14GPT-6 SolOpenAI$1,350
  15. 15GPT-5.6 TerraOpenAI$1,380
  16. 16Gemini 3.1 ProGoogle$1,380
  17. 17DeepSeek V4 ProDeepSeek$1,703
  18. 18Grok 4.7xAI$2,580
  19. 19Qwen3.8 MaxAlibaba$2,580
  20. 20Claude Opus 5.5Anthropic$2,700
  21. 21GPT-5.6 SolOpenAI$2,700
  22. 22Claude Opus 5Anthropic$3,375
  23. 23Claude Fable 5.1Anthropic$6,750
  24. 24GPT-6 AstraOpenAI$6,750

How the LLM Classification Cost Calculator works

cost = (fresh input × input rate + cached input × cache rate + (output + reasoning) × output rate) ÷ 1,000,000

This page uses the same method as the LLM API Cost Calculator, which explains each term and assumption in full.

Worked example

100,000 short texts a day, such as tickets or reviews, each tagged with one label. About 250 tokens of instructions and labels plus the text, and a few tokens back. Prompts this short fall under most providers' minimum cacheable size, so nothing is cached. Priced on GPT-6 Luna through the batch API.

  1. 1Each request: 400 input tokens, 10 output
  2. 2Input costs $0.00002 and output $0.0000025, so $0.0000225 a request
  3. 3× 3,000,000 requests over 30 days = $67.50

$67.50 a month on GPT-6 Luna. The cheapest listed model, Claude Haiku 5.5, would cost $67.50.

Frequently asked questions

Which model is cheapest for classification?

The smallest models: at list rates GPT-6 Luna, GPT-5.6 Luna and Gemini 3.5 Flash-Lite price a short classification request at a small fraction of a cent. Classification is input-heavy with a tiny output, so the input rate matters most. Check accuracy on a labelled sample before choosing on price.

How can I cut classification costs further?

Send several items per request and ask for one label each, which shares the instructions across items. Keep the label list short and ask for the label alone, with no explanation. Run the job through the batch API for half price.

Sources

Related calculators

More in AI, Cloud & Developer