GPT-5.6 vs Qwen 3.8 27B: which is better at resume extraction?
GPT-5.6 leads Qwen 3.8 27B by 1.0 points (84.7 vs 83.8 F1). Cost per 1,000 pages: $74 vs $9.59. Median time per resume: 38 s vs 36 s.
GPT-5.6 wins on 143 of 300 resumes, Qwen 3.8 27B on 117, 40 tied.
84.75
GPT-5.6
A · Vision LLM
83.77
Qwen 3.8 27B
B · Vision LLM
GPT-5.6 is ahead by 0.98 points on these 300 resumes.
- Many entries 83A 84.5 · B 81.6+2.8
- Long bullet lists 101A 84.9 · B 80.8+4.1
- Split across pages 85A 84.9 · B 81.8+3.0
- Big skills list 60A 85.2 · B 82.9+2.3
- Custom sections 105A 82.2 · B 81.0+1.2
- Combined headings 37A 84.1 · B 83.7+0.4
- Promotions 88A 85.0 · B 83.2+1.8
- Partial dates 138A 85.7 · B 83.6+2.1
- Overlapping jobs 69A 83.2 · B 82.4+0.8
- Accented text 13A 86.5 · B 84.9+1.6
Side by side
| GPT-5.6 | Qwen 3.8 27B | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | OpenAI | Alibaba |
| Rank | 6 of 18 | 9 of 18 |
| F1 | 84.7 | 83.8 |
| Precision | 85.9 | 85.4 |
| Recall | 87.2 | 85.5 |
| Cost per 1,000 pages | $74 | $9.59 |
| Median time per resume | 38 s | 36 s |
| Resumes won | 143 | 117 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | GPT-5.6 | Qwen 3.8 27B | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 19 | 71.6 | 78.4 | −6.9 |
| Rotated & warped pages (distractors) | 21 | 73.2 | 78.7 | −5.5 |
| Image only (reading the page) | 28 | 65.4 | 70.5 | −5.1 |
| Hidden text & prompt injections (distractors) | 14 | 77.5 | 82.4 | −4.9 |
| Long (length) | 62 | 77.7 | 73.4 | +4.3 |
| Long bullet lists (extraction challenges) | 101 | 84.9 | 80.8 | +4.1 |
| Busy page (noise) | 18 | 84.0 | 80.4 | +3.7 |
| Content as images (distractors) | 8 | 93.0 | 89.5 | +3.5 |
| Split across pages (extraction challenges) | 85 | 84.9 | 81.8 | +3.0 |
| Resume-builder templates (source) | 10 | 83.7 | 86.7 | −3.0 |
| References (noise) | 55 | 87.9 | 84.9 | +3.0 |
| Many entries (extraction challenges) | 83 | 84.5 | 81.6 | +2.8 |
| Section | Resumes | GPT-5.6 | Qwen 3.8 27B | Gap |
|---|---|---|---|---|
| Contact details | 300 | 94.1 | 93.2 | +0.9 |
| Experience | 300 | 88.4 | 86.2 | +2.3 |
| Education | 297 | 91.9 | 87.2 | +4.7 |
| Summary | 290 | 83.4 | 81.7 | +1.7 |
| Certifications | 255 | 77.9 | 82.7 | −4.9 |
| Awards | 95 | 81.2 | 78.4 | +2.7 |
| Skills | 287 | 93.2 | 92.7 | +0.5 |
| Languages | 168 | 85.9 | 85.4 | +0.5 |
| Custom sections | 212 | 51.2 | 51.2 | +0.0 |
| Interests | 64 | 91.4 | 94.6 | −3.2 |
| Volunteering | 77 | 85.7 | 84.7 | +1.0 |
| Profiles | 37 | 69.4 | 72.5 | −3.2 |
| Projects | 80 | 88.2 | 85.5 | +2.7 |
| Publications | 51 | 77.8 | 77.2 | +0.6 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | GPT-5.6 | Qwen 3.8 27B | Gap |
|---|---|---|---|
| Postdoctoral Researcher: Academic CV, seven pages | 72.3 | 87.0 | −14.7 |
| Lead Respiratory Therapist (Adult ICU): Long resume, 34 entries | 88.7 | 74.2 | +14.6 |
| Senior Controls Engineer: Five custom sections | 86.8 | 73.5 | +13.3 |
| Creative Director: Accented text in a styled CV | 90.8 | 78.3 | +12.4 |
Which to pick
GPT-5.6 is more accurate. Qwen 3.8 27B costs 87% less for 1.0 fewer points.
GPT-5.6 does better on long, long bullet lists and busy page. Qwen 3.8 27B does better on scanned & rotated, rotated & warped pages and image only.
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 GPT-5.4 Nano
- 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 3.5 35B
- GPT-5.6 vs Qwen 3.5 9B
- GPT-5.6 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-5.5
- Qwen 3.8 27B vs GPT-5.4 Nano
- 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