Free tool / build
Fix that broken JSON.
Paste broken output from ChatGPT, Claude or any model and get valid JSON back, with a list of what was fixed.
No sign-up / runs in your browser
From almost-JSON to valid JSON.
Paste broken JSON, or try the example. You will see the fixed version and a list of exactly what was changed.
Repair never invents data. If the output was cut off, it ends where the original ended and open brackets are closed. Always check the result, because a model may have changed the meaning before the formatting broke. To prevent this at the source, use your provider's structured output or JSON mode.
How it works
Three steps, no sign-up.
- 01
Paste the output
Anything meant to be JSON, even with chat around it or a code fence.
- 02
See the fixed version
Valid, indented JSON, and a list of every kind of fix that was made.
- 03
Copy or download
Copy the result, download it, or switch to minified.
Questions people ask
What kinds of mistakes does it fix?
Code fences and chatty text around the data, single quotes, unquoted keys, trailing commas, comments, Python True/False/None, curly quotes, missing commas, raw line breaks inside strings, quotes inside strings and output cut off before it closes.
Is my data sent anywhere?
No. The repair runs in your browser.
Can it fix cut-off JSON?
Yes. It closes open strings, lists and objects where the text stops, and drops a half-written final entry. It does not guess the missing data.
Why does my model return invalid JSON?
Long outputs get truncated, models add explanations, or the prompt does not constrain the format. Use structured outputs or a JSON schema, and validate with a retry.
How do I find out how often this happens?
Tracify traces every model call, so you can see how often outputs fail to parse and which prompts cause it.
More free tools
- Agent log checker
Paste a log or trace. See where it failed, what was slow and what wasted tokens.
- Token counter
Paste text. See the token count and price on GPT, Claude, Gemini and Llama.
- Model price comparison
Every model's price in one table, plus what your app costs per month on each.
- Prompt cost checker
Paste a prompt. Get a shorter, cheaper version and what it costs on each model.
Now measure the real thing.
Send a first trace and see the true cost, speed and failures of every run.