GPT-5.4 Nano vs Qwen 3.8 27B: which is better at resume extraction?
Qwen 3.8 27B leads GPT-5.4 Nano by 21.9 points (83.8 vs 61.9 F1). Cost per 1,000 pages: $2.76 vs $9.59. Median time per resume: 20 s vs 36 s.
GPT-5.4 Nano wins on 25 of 300 resumes, Qwen 3.8 27B on 272, 3 tied.
61.86
GPT-5.4 Nano
A · Vision LLM
83.77
Qwen 3.8 27B
B · Vision LLM
Qwen 3.8 27B is ahead by 21.91 points on these 300 resumes.
- Many entries 83A 59.7 · B 81.6−22.0
- Long bullet lists 101A 61.0 · B 80.8−19.8
- Split across pages 85A 61.7 · B 81.8−20.1
- Big skills list 60A 64.9 · B 82.9−18.0
- Custom sections 105A 59.5 · B 81.0−21.5
- Combined headings 37A 64.6 · B 83.7−19.1
- Promotions 88A 58.7 · B 83.2−24.5
- Partial dates 138A 62.6 · B 83.6−21.0
- Overlapping jobs 69A 58.6 · B 82.4−23.7
- Accented text 13A 55.9 · B 84.9−29.1
Side by side
| GPT-5.4 Nano | Qwen 3.8 27B | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | OpenAI | Alibaba |
| Rank | 17 of 18 | 9 of 18 |
| F1 | 61.9 | 83.8 |
| Precision | 70.1 | 85.4 |
| Recall | 66.3 | 85.5 |
| Cost per 1,000 pages | $2.76 | $9.59 |
| Median time per resume | 20 s | 36 s |
| Resumes won | 25 | 272 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | GPT-5.4 Nano | Qwen 3.8 27B | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 19 | 6.3 | 78.4 | −72.1 |
| Rotated & warped pages (distractors) | 21 | 11.7 | 78.7 | −67.1 |
| Hidden text & prompt injections (distractors) | 14 | 22.0 | 82.4 | −60.4 |
| Image only (reading the page) | 28 | 21.0 | 70.5 | −49.5 |
| Typos kept on purpose (distractors) | 37 | 33.5 | 82.9 | −49.4 |
| Two-column sidebar (layout) | 54 | 47.4 | 78.4 | −31.0 |
| Accented text (extraction challenges) | 13 | 55.9 | 84.9 | −29.1 |
| Content as images (distractors) | 8 | 61.3 | 89.5 | −28.2 |
| Sidebar or grid (reading the page) | 82 | 54.9 | 81.8 | −26.9 |
| Distractor layouts (source) | 18 | 62.2 | 87.6 | −25.4 |
| Medium (length) | 90 | 60.2 | 85.5 | −25.3 |
| Wording traps (distractors) | 16 | 62.8 | 87.6 | −24.8 |
| Section | Resumes | GPT-5.4 Nano | Qwen 3.8 27B | Gap |
|---|---|---|---|---|
| Contact details | 300 | 81.2 | 93.2 | −12.0 |
| Experience | 300 | 63.7 | 86.2 | −22.5 |
| Education | 297 | 62.2 | 87.2 | −25.0 |
| Summary | 290 | 71.7 | 81.7 | −10.0 |
| Certifications | 244 | 53.6 | 82.7 | −29.2 |
| Awards | 102 | 44.9 | 78.4 | −33.5 |
| Skills | 291 | 74.8 | 92.7 | −17.9 |
| Languages | 170 | 69.0 | 85.4 | −16.4 |
| Custom sections | 231 | 26.3 | 51.2 | −24.9 |
| Interests | 66 | 70.0 | 94.6 | −24.6 |
| Volunteering | 80 | 59.9 | 84.7 | −24.8 |
| Profiles | 66 | 23.7 | 72.5 | −48.8 |
| Projects | 81 | 48.8 | 85.5 | −36.7 |
| Publications | 51 | 30.7 | 77.2 | −46.5 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | GPT-5.4 Nano | Qwen 3.8 27B | Gap |
|---|---|---|---|
| Deputy Head of Mission: Rotated scan with hidden identity text | 1.7 | 65.6 | −63.9 |
| Supervisory Human Resources Specialist, GS-0201-15: Resume-builder template | 32.5 | 88.5 | −56.0 |
| Partner: Consulting CV, six pages | 27.8 | 79.7 | −51.9 |
| Head of Digital Marketing: Canva design with a photo | 39.3 | 84.9 | −45.6 |
Which to pick
Qwen 3.8 27B is more accurate. GPT-5.4 Nano costs 71% less for 21.9 fewer points.
Qwen 3.8 27B 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-5.6
- 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 3.5 35B
- GPT-5.4 Nano vs Qwen 3.5 9B
- GPT-5.4 Nano vs GLM 5.3 Flash
Qwen 3.8 27B against every other model
- Qwen 3.8 27B vs LlamaParse Extract Agentic Plus
- Qwen 3.8 27B vs LlamaParse Extract Agentic
- Qwen 3.8 27B vs LlamaParse Extract Cost-Effective
- Qwen 3.8 27B vs Reducto Deep Extract
- Qwen 3.8 27B vs Reducto Extract
- Qwen 3.8 27B vs Extend Max Context
- Qwen 3.8 27B vs Extend Extract
- Qwen 3.8 27B vs GPT-5.6
- Qwen 3.8 27B vs GPT-5.5
- Qwen 3.8 27B vs Claude Opus 4.8
- Qwen 3.8 27B vs Gemini 3.5 Flash
- Qwen 3.8 27B vs Gemma 4 26B
- Qwen 3.8 27B vs Qwen 3.8 Flash
- Qwen 3.8 27B vs Qwen 3.5 35B
- Qwen 3.8 27B vs Qwen 3.5 9B
- Qwen 3.8 27B 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