GPT-5.4 Nano vs GPT-5.6: which is better at resume extraction?
GPT-5.6 leads GPT-5.4 Nano by 22.9 points (84.7 vs 61.9 F1). Cost per 1,000 pages: $2.76 vs $74. Median time per resume: 20 s vs 38 s.
GPT-5.4 Nano wins on 20 of 300 resumes, GPT-5.6 on 278, 2 tied.
61.86
GPT-5.4 Nano
A · Vision LLM
84.75
GPT-5.6
B · Vision LLM
GPT-5.6 is ahead by 22.89 points on these 300 resumes.
- Many entries 83A 59.7 · B 84.5−24.8
- Long bullet lists 101A 61.0 · B 84.9−23.9
- Split across pages 85A 61.7 · B 84.9−23.2
- Big skills list 60A 64.9 · B 85.2−20.3
- Custom sections 105A 59.5 · B 82.2−22.7
- Combined headings 37A 64.6 · B 84.1−19.5
- Promotions 88A 58.7 · B 85.0−26.3
- Partial dates 138A 62.6 · B 85.7−23.1
- Overlapping jobs 69A 58.6 · B 83.2−24.5
- Accented text 13A 55.9 · B 86.5−30.7
Side by side
| GPT-5.4 Nano | GPT-5.6 | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | OpenAI | OpenAI |
| Rank | 17 of 18 | 6 of 18 |
| F1 | 61.9 | 84.7 |
| Precision | 70.1 | 85.9 |
| Recall | 66.3 | 87.2 |
| Cost per 1,000 pages | $2.76 | $74 |
| Median time per resume | 20 s | 38 s |
| Resumes won | 20 | 278 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | GPT-5.4 Nano | GPT-5.6 | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 19 | 6.3 | 71.6 | −65.2 |
| Rotated & warped pages (distractors) | 21 | 11.7 | 73.2 | −61.5 |
| Hidden text & prompt injections (distractors) | 14 | 22.0 | 77.5 | −55.5 |
| Typos kept on purpose (distractors) | 37 | 33.5 | 80.6 | −47.0 |
| Image only (reading the page) | 28 | 21.0 | 65.4 | −44.5 |
| Content as images (distractors) | 8 | 61.3 | 93.0 | −31.7 |
| Two-column sidebar (layout) | 54 | 47.4 | 78.3 | −30.9 |
| Accented text (extraction challenges) | 13 | 55.9 | 86.5 | −30.7 |
| Distractor layouts (source) | 18 | 62.2 | 90.1 | −27.8 |
| Sidebar or grid (reading the page) | 82 | 54.9 | 81.7 | −26.8 |
| Promotions (extraction challenges) | 88 | 58.7 | 85.0 | −26.3 |
| Wording traps (distractors) | 16 | 62.8 | 89.1 | −26.3 |
| Section | Resumes | GPT-5.4 Nano | GPT-5.6 | Gap |
|---|---|---|---|---|
| Contact details | 300 | 81.2 | 94.1 | −12.9 |
| Experience | 300 | 63.7 | 88.4 | −24.7 |
| Education | 297 | 62.2 | 91.9 | −29.7 |
| Summary | 290 | 71.7 | 83.4 | −11.7 |
| Certifications | 244 | 53.6 | 77.9 | −24.3 |
| Awards | 102 | 44.9 | 81.2 | −36.2 |
| Skills | 291 | 74.8 | 93.2 | −18.5 |
| Languages | 170 | 69.0 | 85.9 | −16.8 |
| Custom sections | 231 | 26.3 | 51.2 | −24.9 |
| Interests | 66 | 70.0 | 91.4 | −21.4 |
| Volunteering | 80 | 59.9 | 85.7 | −25.8 |
| Profiles | 66 | 23.7 | 69.4 | −45.6 |
| Projects | 81 | 48.8 | 88.2 | −39.4 |
| Publications | 51 | 30.7 | 77.8 | −47.1 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | GPT-5.4 Nano | GPT-5.6 | Gap |
|---|---|---|---|
| Deputy Head of Mission: Rotated scan with hidden identity text | 1.7 | 76.2 | −74.5 |
| Partner: Consulting CV, six pages | 27.8 | 79.9 | −52.1 |
| Head of Digital Marketing: Canva design with a photo | 39.3 | 86.7 | −47.5 |
| Supervisory Human Resources Specialist, GS-0201-15: Resume-builder template | 32.5 | 78.6 | −46.2 |
Which to pick
GPT-5.6 is more accurate. GPT-5.4 Nano costs 96% less for 22.9 fewer points.
GPT-5.6 does better on scanned & rotated, rotated & warped pages and hidden text & prompt injections.
GPT-5.4 Nano against every other model
- GPT-5.4 Nano vs LlamaParse Extract Agentic Plus
- GPT-5.4 Nano vs LlamaParse Extract Agentic
- GPT-5.4 Nano vs LlamaParse Extract Cost-Effective
- GPT-5.4 Nano vs Reducto Deep Extract
- GPT-5.4 Nano vs Reducto Extract
- GPT-5.4 Nano vs Extend Max Context
- GPT-5.4 Nano vs Extend Extract
- GPT-5.4 Nano vs GPT-5.5
- GPT-5.4 Nano vs Claude Opus 4.8
- GPT-5.4 Nano vs Gemini 3.5 Flash
- GPT-5.4 Nano vs Gemma 4 26B
- GPT-5.4 Nano vs Qwen 3.8 Flash
- GPT-5.4 Nano vs Qwen 3.8 27B
- GPT-5.4 Nano vs Qwen 3.5 35B
- GPT-5.4 Nano vs Qwen 3.5 9B
- GPT-5.4 Nano vs GLM 5.3 Flash
GPT-5.6 against every other model
- GPT-5.6 vs LlamaParse Extract Agentic Plus
- GPT-5.6 vs LlamaParse Extract Agentic
- GPT-5.6 vs LlamaParse Extract Cost-Effective
- GPT-5.6 vs Reducto Deep Extract
- GPT-5.6 vs Reducto Extract
- GPT-5.6 vs Extend Max Context
- GPT-5.6 vs Extend Extract
- GPT-5.6 vs GPT-5.5
- GPT-5.6 vs Claude Opus 4.8
- GPT-5.6 vs Gemini 3.5 Flash
- GPT-5.6 vs Gemma 4 26B
- GPT-5.6 vs Qwen 3.8 Flash
- GPT-5.6 vs Qwen 3.8 27B
- GPT-5.6 vs Qwen 3.5 35B
- GPT-5.6 vs Qwen 3.5 9B
- GPT-5.6 vs GLM 5.3 Flash
Popular comparisons
- LlamaParse Extract Agentic Plus vs Reducto Deep Extract
- Extend Max Context vs Reducto Deep Extract
- Extend Max Context vs LlamaParse Extract Agentic Plus
- Claude Opus 4.8 vs GPT-5.6
- Gemini 3.5 Flash vs GPT-5.6
- Claude Opus 4.8 vs Gemini 3.5 Flash
- LlamaParse Extract Agentic vs LlamaParse Extract Agentic Plus
- Reducto Deep Extract vs Reducto Extract
- Gemma 4 26B vs Qwen 3.8 27B
- Document APIs vs vision LLMs