Skip to content

Which resume extraction model handles handwritten resumes best?

9 of the 300 resumes are handwritten resumes. The typical system scores 31.7 points lower on them. Best: LlamaParse Extract Agentic Plus, 67.8 F1.

Resumes written by hand and scanned or photographed. There is no text layer: every word has to be read from the handwriting. Best document API: LlamaParse Extract Agentic Plus (67.8). Best vision LLM: Claude Opus 4.8 (54.7). GLM 5.3 Flash ranks 14 places higher here than overall.

Every system on handwritten resumes

F1 on the 9 handwritten resumes, against each system’s F1 on all 300 resumes. Cost and time are for all resumes.

Ranking on handwritten resumes
ModelF1 hereF1 overallGapPer 1,000 pagesMedian time
LlamaParse Extract Agentic Plus67.887.2−19.3$5470 s
LlamaParse Extract Agentic63.585.5−22.1$2355 s
Claude Opus 4.854.785.5−30.8$9031 s
GLM 5.3 Flash54.659.2−4.6$0.8046 s
Qwen 3.8 Flash53.584.1−30.6$2.98122 s
GPT-5.653.284.7−31.5$7438 s
Qwen 3.5 35B52.777.9−25.2$3.9178 s
Extend Extract52.581.0−28.5$1231 s
Qwen 3.8 27B51.683.8−32.2$9.5936 s
GPT-5.4 Nano51.361.9−10.5$2.7620 s
GPT-5.551.083.5−32.5$8748 s
Qwen 3.5 9B50.676.0−25.4$1.31138 s
Gemini 3.5 Flash50.385.5−35.2$3543 s
Gemma 4 26B50.362.8−12.5$0.5629 s
LlamaParse Extract Cost-Effective49.683.5−33.9$8.1829 s
Reducto Deep Extract49.485.6−36.2$4065 s
Extend Max Context46.584.7−38.2$5043 s
Reducto Extract39.481.7−42.3$2016 s

What to do about it

Every system struggles here. Use the best one below, or send these to a person.

Published examples

2 public resumes of this kind.

Other hard resumes

Frequently asked questions (FAQs)

Which resume extraction model is best for handwritten resumes?

LlamaParse Extract Agentic Plus, at 67.8 F1.

How hard are handwritten resumes for parsers?

The typical model scores 31.7 points lower on them than on other resumes.

How many handwritten resumes are in ResumeBench?

9 of the 300.