GPT-5.6 vs Qwen 3.8 Flash: which is better at resume extraction?
GPT-5.6 leads Qwen 3.8 Flash by 0.7 points (84.7 vs 84.1 F1). Cost per 1,000 pages: $74 vs $2.98. Median time per resume: 38 s vs 122 s.
GPT-5.6 wins on 157 of 298 resumes, Qwen 3.8 Flash on 102, 39 tied.
84.80
GPT-5.6
A · Vision LLM
84.08
Qwen 3.8 Flash
B · Vision LLM
GPT-5.6 is ahead by 0.72 points on these 298 resumes.
- Many entries 83A 84.5 · B 82.1+2.4
- Long bullet lists 101A 84.9 · B 82.1+2.9
- Split across pages 85A 84.9 · B 82.6+2.2
- Big skills list 60A 85.2 · B 83.5+1.7
- Custom sections 105A 82.2 · B 80.8+1.3
- Combined headings 37A 84.1 · B 82.6+1.4
- Promotions 86A 85.2 · B 83.8+1.4
- Partial dates 138A 85.7 · B 83.3+2.4
- Overlapping jobs 68A 83.2 · B 81.4+1.8
- Accented text 13A 86.5 · B 86.9−0.3
Side by side
| GPT-5.6 | Qwen 3.8 Flash | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | OpenAI | Alibaba |
| Rank | 6 of 18 | 8 of 18 |
| F1 | 84.7 | 84.1 |
| Precision | 85.9 | 85.4 |
| Recall | 87.2 | 86.2 |
| Cost per 1,000 pages | $74 | $2.98 |
| Median time per resume | 38 s | 122 s |
| Resumes won | 157 | 102 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | GPT-5.6 | Qwen 3.8 Flash | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 17 | 70.9 | 86.3 | −15.4 |
| Rotated & warped pages (distractors) | 19 | 72.8 | 86.0 | −13.2 |
| Hidden text & prompt injections (distractors) | 13 | 77.2 | 89.0 | −11.8 |
| Image only (reading the page) | 26 | 64.6 | 74.7 | −10.2 |
| Typos kept on purpose (distractors) | 35 | 80.8 | 85.7 | −5.0 |
| Distractor layouts (source) | 18 | 90.1 | 85.1 | +4.9 |
| Content as images (distractors) | 8 | 93.0 | 88.3 | +4.8 |
| Busy page (noise) | 18 | 84.0 | 79.8 | +4.2 |
| Two-column sidebar (layout) | 52 | 78.4 | 81.6 | −3.2 |
| Wording traps (distractors) | 16 | 89.1 | 86.0 | +3.1 |
| Long bullet lists (extraction challenges) | 101 | 84.9 | 82.1 | +2.9 |
| Long (length) | 62 | 77.7 | 75.0 | +2.7 |
| Section | Resumes | GPT-5.6 | Qwen 3.8 Flash | Gap |
|---|---|---|---|---|
| Contact details | 298 | 94.1 | 93.6 | +0.5 |
| Experience | 298 | 88.5 | 88.0 | +0.6 |
| Education | 295 | 91.8 | 90.1 | +1.7 |
| Summary | 288 | 83.7 | 81.3 | +2.4 |
| Certifications | 254 | 77.9 | 82.2 | −4.3 |
| Awards | 93 | 80.8 | 75.5 | +5.2 |
| Skills | 285 | 93.3 | 90.0 | +3.3 |
| Languages | 166 | 85.7 | 84.6 | +1.1 |
| Custom sections | 210 | 51.4 | 53.9 | −2.5 |
| Interests | 63 | 91.2 | 91.1 | +0.1 |
| Volunteering | 76 | 85.9 | 83.9 | +2.0 |
| Profiles | 37 | 69.4 | 73.3 | −4.0 |
| Projects | 80 | 88.2 | 83.7 | +4.5 |
| Publications | 51 | 77.8 | 81.8 | −4.1 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | GPT-5.6 | Qwen 3.8 Flash | Gap |
|---|---|---|---|
| Creative Director: Accented text in a styled CV | 90.8 | 76.8 | +14.0 |
| Senior Customer Service Representative: Busy Canva page | 83.1 | 94.9 | −11.9 |
| Deputy Head of Mission: Rotated scan with hidden identity text | 76.2 | 88.0 | −11.8 |
| Lead Solutions Architect: Indian format with CGPA | 83.8 | 72.7 | +11.1 |
Which to pick
GPT-5.6 is more accurate. Qwen 3.8 Flash costs 96% less for 0.7 fewer points.
GPT-5.6 does better on distractor layouts, content as images and busy page. Qwen 3.8 Flash does better on scanned & rotated, rotated & warped pages and hidden text & prompt injections.
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 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
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-5.5
- Qwen 3.8 Flash vs GPT-5.4 Nano
- 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