OCR For construction documents does not work, we fixed it
Traditional OCR often fails with complex technical drawings. This new solution enables automatic door detection on architectural floor plans via a specialized API, optimizing the quantity takeoff process.
Door detection processing time (free tier)
This applies only to door detection (POST /v1/drawings/detection/doors).
Free-tier door detection typically completes in 2–4 minutes per job. Processing time scales mainly with page count and drawing complexity — a dense multi-page set takes longer than a single sheet.
Need it faster? Pro and Enterprise plans run detection on dedicated GPU infrastructure.
/v1/drawings/detection/doors
Detect doors in architectural floor-plan PDFs. Accepts a previously uploaded document_id, enqueues inference, and returns a job you poll for results. Detections are returned as bounding boxes in PDF coordinate space.
Request
Auth via X-API-Key header. Body is JSON — this endpoint does not accept file uploads directly.
Code examples
Response — 202 Accepted
The job is enqueued immediately. Poll GET /v1/jobs/{job_id} until status is complete or failed.
Result shape
When status === "complete" and model === "door-detector", the result field on the job contains:
Credits & rate limits
Rate-limit 429s include retry_after_seconds in the body. Quota 429s and rate-limit 429s return the same status code — check the error body to distinguish them.
Source: Hacker News
