> ## Documentation Index
> Fetch the complete documentation index at: https://docs.parsetx.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Error Handling

> Handling HTTP errors and partial batch failures.

ParseTx uses conventional HTTP response codes.

## HTTP Error Codes

| Code  | Meaning        | Resolution                                                                         |
| :---- | :------------- | :--------------------------------------------------------------------------------- |
| `400` | Bad Request    | You submitted invalid JSON or violated the schema limits (e.g. >10 items on sync). |
| `401` | Unauthorized   | Your API key is missing or invalid. Check the dashboard.                           |
| `429` | Quota Exceeded | You have hit your token limit or concurrency limit.                                |
| `500` | Server Error   | Something went wrong on our end.                                                   |

## Partial Batch Failures

Because we process transactions in batches, a single bad transaction does NOT fail the entire request.

Instead, the overall response returns `200 OK`, but individual items inside the `results` array may have a `status` of `retry` or `rejected`.

* **`rejected`**: The input string was garbage (e.g., all numbers, or just "123"). It was rejected by our sanitization pipeline. Do not retry it.
* **`retry`**: Our live LLM inference engine timed out. You should retry this specific string later.