Test Launch

Improved document extraction accuracy
We’ve tightened OCR handling for low-contrast and scanned documents, improving extraction consistency across edge cases.

Faster API responses
Reduced latency on several core endpoints, especially under higher concurrency.

More detailed error messages
You’ll now see clearer, more actionable error responses when requests fail.