Side-by-Side AI Model Cost Comparison

Select 2 to 4 models and enter your workload to see a detailed cost comparison. Last updated:

Which AI model is cheapest for my workload? Pick up to four models from the dropdowns below, enter your expected input and output token counts, and this calculator shows you the cost of each model side by side. It displays input rates, output rates, cached rates, per-workload costs, and highlights the cheapest option with percentage savings over the most expensive pick.

Select Models to Compare (2-4)

Workload

100,000 tokens (~75,000 words)
50,000 tokens (~37,500 words)

Select at least 2 models above to see a side-by-side comparison.

How does the model comparison work?

  1. Select 2 to 4 models — choose the models you're considering from the dropdowns. All models from our pricing database are available.
  2. Enter your workload — set the input and output token counts for a typical request or batch job.
  3. Compare the table — see per-million rates, computed costs, cached-input costs, and which model is cheapest.
  4. Check the verdict — the summary shows the cheapest model and the dollar and percentage savings vs the most expensive option.

Methodology

For each selected model, the calculator applies the standard cost formula:

cost = (input_tokens / 1,000,000 × input_rate) + (output_tokens / 1,000,000 × output_rate)

The "With 100% Cache" row shows the cost if all input tokens hit the provider's cached-input rate (where available). This is a best-case scenario — real cache hit rates depend on your prompt structure and the provider's caching implementation.

The "vs Cheapest" row shows how much more each model costs compared to the cheapest selected model for the same workload. All rates are sourced from official provider pricing pages, updated daily.

Need AI for content at scale? Writesonic lets you generate blog posts, ad copy, and product descriptions with a simple interface — no API integration needed.