Quick start
-
Paste output
Fences with markdown, json, etc. are stripped.
-
Copy cleaned text
Keeps body text; removes formatting noise only.
LLM Response Cleaner
Models often wrap Markdown in fences with extra blank lines. Paste and clean in one step for storage or reuse.
Read the full guide: The Complete Guide to OpenAI Message Formatting & LLM Data Processing →
Privacy: processed locally, never uploaded.
↓ Paste in the input area below to see results instantly
Raw LLM output
Strip Markdown fences, extra blank lines, and trailing whitespace.
Cleaned text
Notes
What it does
Removes ``` fences; collapses 3+ newlines to 2; trims trailing spaces per line. Does not rewrite content.
Models often wrap Markdown in fences with extra blank lines. Paste and clean in one step for storage or reuse.
What it does not do
No Markdown parsing, translation, or summarization; light text hygiene only.
Features and use cases
Strip Markdown fences, extra blank lines, and common prefixes from LLM output for clean text or JSON.
Use before pipelines, docs paste, or as a pre-step to llm-json-extractor.
Typical Workflow
When copying responses from models like ChatGPT, they often contain extra code fences and blank lines. Paste this text into the cleaner tool, click the button, and get clean content. Ideal for storing model outputs in databases, publishing to websites, or further code processing.
The cleaned text can be directly used in Markdown editors or CMS systems without manually removing fences. For developers, the plain text is easier to process with scripts, such as batch keyword replacement or content structure analysis.
Examples
Fenced block
Input
```markdown\n# Hi\n```
Output
# Hi
FAQ
Remove code block content?
Only fence lines are removed; inner content stays.
Can the tool handle nested code fences?
No. The tool removes all ``` fences, including nested ones. To preserve nested structures, consider manual handling or specialized regex tools.