GPT-5.4 Nano vs Qwen 3.8 Flash: which is better at resume extraction?
Qwen 3.8 Flash leads GPT-5.4 Nano by 22.2 points (84.1 vs 61.9 F1). Cost per 1,000 pages: $2.76 vs $2.98. Median time per resume: 20 s vs 122 s.
GPT-5.4 Nano wins on 33 of 298 resumes, Qwen 3.8 Flash on 264, 1 tied.
62.25
GPT-5.4 Nano
A · Vision LLM
84.08
Qwen 3.8 Flash
B · Vision LLM
Qwen 3.8 Flash is ahead by 21.83 points on these 298 resumes.
- Many entries 83A 59.7 · B 82.1−22.4
- Long bullet lists 101A 61.0 · B 82.1−21.0
- Split across pages 85A 61.7 · B 82.6−20.9
- Big skills list 60A 64.9 · B 83.5−18.6
- Custom sections 105A 59.5 · B 80.8−21.3
- Combined headings 37A 64.6 · B 82.6−18.0
- Promotions 86A 60.0 · B 83.8−23.8
- Partial dates 138A 62.6 · B 83.3−20.7
- Overlapping jobs 68A 59.4 · B 81.4−21.9
- Accented text 13A 55.9 · B 86.9−31.0
Side by side
| GPT-5.4 Nano | Qwen 3.8 Flash | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | OpenAI | Alibaba |
| Rank | 17 of 18 | 8 of 18 |
| F1 | 61.9 | 84.1 |
| Precision | 70.1 | 85.4 |
| Recall | 66.3 | 86.2 |
| Cost per 1,000 pages | $2.76 | $2.98 |
| Median time per resume | 20 s | 122 s |
| Resumes won | 33 | 264 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | GPT-5.4 Nano | Qwen 3.8 Flash | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 17 | 6.6 | 86.3 | −79.7 |
| Rotated & warped pages (distractors) | 19 | 12.5 | 86.0 | −73.5 |
| Hidden text & prompt injections (distractors) | 13 | 23.4 | 89.0 | −65.6 |
| Image only (reading the page) | 26 | 22.3 | 74.7 | −52.4 |
| Typos kept on purpose (distractors) | 35 | 35.2 | 85.7 | −50.5 |
| Two-column sidebar (layout) | 52 | 49.1 | 81.6 | −32.5 |
| Accented text (extraction challenges) | 13 | 55.9 | 86.9 | −31.0 |
| Sidebar or grid (reading the page) | 80 | 56.1 | 83.8 | −27.7 |
| Content as images (distractors) | 8 | 61.3 | 88.3 | −27.0 |
| Photo (reading the page) | 44 | 60.6 | 86.3 | −25.7 |
| Medium (length) | 90 | 60.2 | 85.3 | −25.1 |
| Expert-built (Canva / LaTeX) (source) | 32 | 60.3 | 84.4 | −24.1 |
| Section | Resumes | GPT-5.4 Nano | Qwen 3.8 Flash | Gap |
|---|---|---|---|---|
| Contact details | 298 | 81.6 | 93.6 | −12.1 |
| Experience | 298 | 64.1 | 88.0 | −23.8 |
| Education | 295 | 62.6 | 90.1 | −27.5 |
| Summary | 288 | 72.2 | 81.3 | −9.0 |
| Certifications | 243 | 53.8 | 82.2 | −28.4 |
| Awards | 100 | 45.8 | 75.5 | −29.7 |
| Skills | 289 | 75.3 | 90.0 | −14.7 |
| Languages | 168 | 69.7 | 84.6 | −14.9 |
| Custom sections | 229 | 26.6 | 53.9 | −27.3 |
| Interests | 65 | 71.1 | 91.1 | −20.1 |
| Volunteering | 79 | 60.6 | 83.9 | −23.3 |
| Profiles | 66 | 23.7 | 73.3 | −49.6 |
| Projects | 81 | 48.8 | 83.7 | −34.9 |
| Publications | 51 | 30.7 | 81.8 | −51.1 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | GPT-5.4 Nano | Qwen 3.8 Flash | Gap |
|---|---|---|---|
| Deputy Head of Mission: Rotated scan with hidden identity text | 1.7 | 88.0 | −86.3 |
| Supervisory Human Resources Specialist, GS-0201-15: Resume-builder template | 32.5 | 83.8 | −51.3 |
| Partner: Consulting CV, six pages | 27.8 | 74.8 | −47.0 |
| Head of Digital Marketing: Canva design with a photo | 39.3 | 85.1 | −45.8 |
Which to pick
Qwen 3.8 Flash is more accurate. GPT-5.4 Nano costs 7% less for 22.2 fewer points.
Qwen 3.8 Flash 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 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
Qwen 3.8 Flash against every other model
- Qwen 3.8 Flash vs LlamaParse Extract Agentic Plus
- Qwen 3.8 Flash vs LlamaParse Extract Agentic
- Qwen 3.8 Flash vs LlamaParse Extract Cost-Effective
- Qwen 3.8 Flash vs Reducto Deep Extract
- Qwen 3.8 Flash vs Reducto Extract
- Qwen 3.8 Flash vs Extend Max Context
- Qwen 3.8 Flash vs Extend Extract
- Qwen 3.8 Flash vs GPT-5.6
- Qwen 3.8 Flash vs GPT-5.5
- Qwen 3.8 Flash vs Claude Opus 4.8
- Qwen 3.8 Flash vs Gemini 3.5 Flash
- Qwen 3.8 Flash vs Gemma 4 26B
- Qwen 3.8 Flash vs Qwen 3.8 27B
- Qwen 3.8 Flash vs Qwen 3.5 35B
- Qwen 3.8 Flash vs Qwen 3.5 9B
- Qwen 3.8 Flash 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