Skip to content

Which resume extraction model handles rotated and warped scans best?

21 of the 300 resumes are rotated and warped scans. The typical system scores 0.4 points higher on them. Best: LlamaParse Extract Agentic Plus, 92.5 F1. LlamaParse Extract Agentic is within two points for $23 per 1,000 pages.

Image-only scans with pages at different rotations, warped pages and photos of printouts. Best document API: LlamaParse Extract Agentic Plus (92.5). Best vision LLM: Gemini 3.5 Flash (87.4). LlamaParse Extract Cost-Effective ranks 6 places higher here than overall.

Every system on rotated and warped scans

F1 on the 21 rotated and warped scans, against each system’s F1 on all 300 resumes. Cost and time are for all resumes.

Ranking on rotated and warped scans
ModelF1 hereF1 overallGapPer 1,000 pagesMedian time
LlamaParse Extract Agentic Plus92.587.2+5.3$5470 s
LlamaParse Extract Agentic90.985.5+5.4$2355 s
Extend Max Context90.884.7+6.0$5043 s
LlamaParse Extract Cost-Effective89.683.5+6.1$8.1829 s
Reducto Deep Extract89.385.6+3.7$4065 s
Reducto Extract89.281.7+7.5$2016 s
Gemini 3.5 Flash87.485.5+2.0$3543 s
Claude Opus 4.887.485.5+1.9$9031 s
Qwen 3.8 Flash86.084.1+1.9$2.98122 s
Extend Extract79.881.0−1.1$1231 s
Qwen 3.8 27B78.783.8−5.0$9.5936 s
Qwen 3.5 35B76.577.9−1.4$3.9178 s
GPT-5.673.284.7−11.5$7438 s
Qwen 3.5 9B66.576.0−9.6$1.31138 s
GPT-5.563.483.5−20.1$8748 s
GLM 5.3 Flash50.359.2−9.0$0.8046 s
Gemma 4 26B43.462.8−19.4$0.5629 s
GPT-5.4 Nano11.761.9−50.2$2.7620 s

What to do about it

If you get phone photos or bad scans, rank parsers on this table.

Published examples

2 public resumes of this kind.

Other hard resumes

Frequently asked questions (FAQs)

Which resume extraction model is best for rotated and warped scans?

LlamaParse Extract Agentic Plus, at 92.5 F1. LlamaParse Extract Agentic is within two points for $23 per 1,000 pages.

How hard are rotated and warped scans for parsers?

The typical model scores 0.4 points higher on them than on other resumes.

How many rotated and warped scans are in ResumeBench?

21 of the 300.