AI Prompt Diff

Runs in browser

Diff two AI prompts — see changes, token delta, and cost impact by model.

Compare two versions of an AI prompt side by side. See added and removed lines highlighted, with token count delta, cost delta per model, and a copy-as-markdown diff output.

AI Prompt Diff tool

17 tokens · $0.00009
45 tokens · $0.00022
+1 lines added1 lines removed+28 tokens+$0.00014 cost/call+111 chars

Diff

You are a helpful assistant. Answer the user's questions concisely.
+You are an expert assistant with deep knowledge across many domains. Answer the user's questions concisely and accurately. When unsure, say so. Always cite sources when possible.

🔒 Runs in your browser · No uploads · Your data never leaves your device

How to use

  1. Enter both prompts

    Paste your original prompt in v1 and the updated version in v2. The diff updates automatically.

  2. Select a model

    Choose the LLM you're using to see the cost impact of the change in $/call.

  3. Review the diff

    Green lines (+) were added in v2. Red lines (−) were removed. Unchanged lines appear in gray.

  4. Copy as markdown

    Click 'Copy as markdown' to get the diff with +/- prefixes for sharing in GitHub issues, Notion, or Slack.

Common use cases

  • Reviewing prompt changes before deploymentDiff the old and new system prompt to confirm what changed and see the token cost impact before shipping.
  • Auditing prompt revisionsCopy the markdown diff to share with teammates in GitHub or Notion for review of prompt changes.

Examples

  • Adding an expert persona

    Compare a basic 'helpful assistant' prompt to a detailed expert persona

    Output
    +4 lines added, +18 tokens, +$0.00009 per call (GPT-4o)

Frequently asked questions

How are tokens calculated?
Token counts use an approximation of ~1 token per 4 English characters. Actual API token counts may vary by ±5% depending on the model's tokenizer.
Are my prompts sent anywhere?
No. All processing happens in your browser. Your prompt text never leaves your device.
What does the cost delta mean?
The cost delta shows how much more or less you'll pay per API call with the new prompt, based on the input pricing of the selected model.

Key concepts

Prompt diff
A line-by-line comparison of two prompt versions, showing additions (green) and removals (red).
Token delta
The change in token count between two prompt versions, used to estimate the cost impact of a prompt edit.

You might find these useful too.

More ai tools