ModelMint

What will your LLM bill actually be?

Describe what you're building in plain terms and see what it would cost per month on the models people are actually choosing between.

What are you building?

Your workload

How much do you send each time?

How long is each answer?

Does every request start with the same text?

The same instructions every time, like a system prompt.

Providers discount repeated text (prompt caching) only when the repeated part is long enough, from 512 to 4,096 tokens depending on the model, and was sent recently.

No caching saving on GPT-6 Luna: the repeated part is 800 tokens, and it only caches 1,024 or more. All input is priced at the normal rate.

I know my exact token counts

Same text % is the share of input tokens that is identical on every request.

Must support

A support chatbot handling 2,000 chats a day costs between $16.50/mo on GPT-6 Luna and $1,650/mo on GPT-6 Astra.

Already using a model?

List prices, 30-day month. An estimate, not a quote.

Select up to 5 models to compare side by side

Flagship

Most capable, most expensive
  1. Gemini Pro
    Google Gemini1M context

    No caching discount assumed: caching rules not verified

    $348/mo
    $0.005800 / chat
  2. Claude Opus 5.5
    Anthropic1M context
    $501/mo
    $0.008344 / chat
  3. Claude Fable 5.1
    Anthropic1M context
    $1,241/mo
    $0.0207 / chat
  4. GPT-6 Astra
    OpenAI922K context

    No caching discount: needs 1,024+ repeated tokens

    $1,650/mo
    $0.0275 / chat

Balanced

Everyday workhorses
  1. Gemini Flash
    Google Gemini1M context

    No caching discount: needs 4,096+ repeated tokens

    $124/mo
    $0.002063 / chat
  2. DeepSeek V4 Pro
    DeepSeek1M context

    No caching discount assumed: caching rules not verified

    Peak-hour rates. Off-peak requests cost half.

    $194/mo
    $0.003234 / chat
  3. Mistral Medium
    Mistral262K context

    No caching discount: no cached-input price

    $248/mo
    $0.004125 / chat
  4. Grok 4.7
    xAI500K context

    No caching discount assumed: caching rules not verified

    $294/mo
    $0.004900 / chat
  5. Claude Sonnet 5
    Anthropic1M context

    No caching discount: needs 1,024+ repeated tokens

    $330/mo
    $0.005500 / chat
  6. GPT-6 Sol
    OpenAI922K context

    No caching discount: needs 1,024+ repeated tokens

    $330/mo
    $0.005500 / chat

Fast & cheap

Small models for high volume
  1. GPT-6 LunaLowest
    OpenAI922K context

    No caching discount: needs 1,024+ repeated tokens

    $16.50/mo
    $0.000275 / chat
  2. Mistral Small
    Mistral262K context

    No caching discount: no cached-input price

    $23.40/mo
    $0.000390 / chat
  3. DeepSeek Flash
    DeepSeek1M context

    No caching discount assumed: caching rules not verified

    Peak-hour rates. Off-peak requests cost half.

    $46.80/mo
    $0.000780 / chat
  4. Gemini Flash-Lite
    Google Gemini1M context

    No caching discount assumed: caching rules not verified

    $58.50/mo
    $0.000975 / chat
  5. Grok Code Fast
    xAI256K context

    No caching discount assumed: caching rules not verified

    $138/mo
    $0.002300 / chat
  6. Claude Haiku 4.5
    Anthropic200K context

    No caching discount: needs 4,096+ repeated tokens

    $165/mo
    $0.002750 / chat

Cheaper alternatives

The cheapest models with the features you need, ranked by price only. ModelMint doesn't measure quality, so try them on your own task before switching.

  1. gpt-5-nanoOpenAI · 272K context
    $9.60/mo
  2. ministral-3b-latestMistral · 131K context
    $12.90/mo
  3. devstral-small-latestMistral · 256K context
    $14.70/mo
  4. mistral-smallMistral · 32K context
    $14.70/mo
  5. gemini-2.5-flash-liteGoogle Gemini · 1M context
    $15.60/mo
Browse all models in the directory →