Gemma 4 26B vs Qwen 3.5 35B: which is better at resume extraction?
Qwen 3.5 35B leads Gemma 4 26B by 15.1 points (77.9 vs 62.8 F1). Cost per 1,000 pages: $0.56 vs $3.91. Median time per resume: 29 s vs 78 s.
Gemma 4 26B wins on 54 of 300 resumes, Qwen 3.5 35B on 243, 3 tied.
62.82
Gemma 4 26B
A · Vision LLM
77.95
Qwen 3.5 35B
B · Vision LLM
Qwen 3.5 35B is ahead by 15.13 points on these 300 resumes.
- Many entries 83A 58.0 · B 78.9−20.9
- Long bullet lists 101A 57.4 · B 76.3−18.9
- Split across pages 85A 60.0 · B 76.5−16.5
- Big skills list 60A 61.9 · B 76.6−14.7
- Custom sections 105A 58.9 · B 77.0−18.1
- Combined headings 37A 60.9 · B 74.6−13.7
- Promotions 88A 59.5 · B 76.7−17.2
- Partial dates 138A 62.9 · B 78.9−16.0
- Overlapping jobs 69A 57.5 · B 76.2−18.7
- Accented text 13A 59.3 · B 80.4−21.2
Side by side
| Gemma 4 26B | Qwen 3.5 35B | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | Alibaba | |
| Rank | 16 of 18 | 14 of 18 |
| F1 | 62.8 | 77.9 |
| Precision | 69.1 | 78.8 |
| Recall | 66.2 | 83.7 |
| Cost per 1,000 pages | $0.56 | $3.91 |
| Median time per resume | 29 s | 78 s |
| Resumes won | 54 | 243 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | Gemma 4 26B | Qwen 3.5 35B | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 19 | 40.7 | 75.3 | −34.6 |
| Rotated & warped pages (distractors) | 21 | 43.4 | 76.5 | −33.1 |
| Hidden text & prompt injections (distractors) | 14 | 48.6 | 79.2 | −30.6 |
| Typos kept on purpose (distractors) | 37 | 51.3 | 80.4 | −29.1 |
| Content as images (distractors) | 8 | 65.6 | 90.3 | −24.7 |
| Image only (reading the page) | 28 | 45.1 | 69.7 | −24.6 |
| Wording traps (distractors) | 16 | 60.6 | 84.8 | −24.2 |
| Distractor layouts (source) | 18 | 62.4 | 85.7 | −23.3 |
| Accented text (extraction challenges) | 13 | 59.3 | 80.4 | −21.2 |
| Many entries (extraction challenges) | 83 | 58.0 | 78.9 | −20.9 |
| Resume-builder templates (source) | 10 | 58.6 | 79.1 | −20.5 |
| References (noise) | 55 | 62.5 | 81.8 | −19.2 |
| Section | Resumes | Gemma 4 26B | Qwen 3.5 35B | Gap |
|---|---|---|---|---|
| Contact details | 300 | 84.6 | 92.1 | −7.5 |
| Experience | 300 | 56.0 | 79.3 | −23.3 |
| Education | 297 | 68.0 | 81.8 | −13.8 |
| Summary | 290 | 35.5 | 80.4 | −44.9 |
| Certifications | 244 | 63.0 | 79.1 | −16.1 |
| Awards | 110 | 55.7 | 61.9 | −6.2 |
| Skills | 290 | 81.9 | 87.8 | −5.9 |
| Languages | 171 | 80.9 | 79.4 | +1.5 |
| Custom sections | 223 | 32.3 | 40.2 | −7.9 |
| Interests | 67 | 71.8 | 89.7 | −17.9 |
| Volunteering | 78 | 59.4 | 83.1 | −23.7 |
| Profiles | 37 | 57.2 | 65.8 | −8.6 |
| Projects | 83 | 55.0 | 76.7 | −21.8 |
| Publications | 53 | 54.5 | 75.0 | −20.5 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | Gemma 4 26B | Qwen 3.5 35B | Gap |
|---|---|---|---|
| Supervisory Human Resources Specialist, GS-0201-15: Resume-builder template | 29.4 | 82.8 | −53.4 |
| Deputy Head of Mission: Rotated scan with hidden identity text | 34.1 | 79.6 | −45.6 |
| Creative Director: Accented text in a styled CV | 39.3 | 79.3 | −40.1 |
| Head of Developer Experience (Internal title: Jedi Master of Analytics) : Unusual job titles | 53.7 | 90.6 | −36.9 |
Which to pick
Qwen 3.5 35B is more accurate. Gemma 4 26B costs 86% less for 15.1 fewer points.
Qwen 3.5 35B does better on scanned & rotated, rotated & warped pages and hidden text & prompt injections.
Gemma 4 26B against every other model
- Gemma 4 26B vs LlamaParse Extract Agentic Plus
- Gemma 4 26B vs LlamaParse Extract Agentic
- Gemma 4 26B vs LlamaParse Extract Cost-Effective
- Gemma 4 26B vs Reducto Deep Extract
- Gemma 4 26B vs Reducto Extract
- Gemma 4 26B vs Extend Max Context
- Gemma 4 26B vs Extend Extract
- Gemma 4 26B vs GPT-5.6
- Gemma 4 26B vs GPT-5.5
- Gemma 4 26B vs GPT-5.4 Nano
- Gemma 4 26B vs Claude Opus 4.8
- Gemma 4 26B vs Gemini 3.5 Flash
- Gemma 4 26B vs Qwen 3.8 Flash
- Gemma 4 26B vs Qwen 3.8 27B
- Gemma 4 26B vs Qwen 3.5 9B
- Gemma 4 26B vs GLM 5.3 Flash
Qwen 3.5 35B against every other model
- Qwen 3.5 35B vs LlamaParse Extract Agentic Plus
- Qwen 3.5 35B vs LlamaParse Extract Agentic
- Qwen 3.5 35B vs LlamaParse Extract Cost-Effective
- Qwen 3.5 35B vs Reducto Deep Extract
- Qwen 3.5 35B vs Reducto Extract
- Qwen 3.5 35B vs Extend Max Context
- Qwen 3.5 35B vs Extend Extract
- Qwen 3.5 35B vs GPT-5.6
- Qwen 3.5 35B vs GPT-5.5
- Qwen 3.5 35B vs GPT-5.4 Nano
- Qwen 3.5 35B vs Claude Opus 4.8
- Qwen 3.5 35B vs Gemini 3.5 Flash
- Qwen 3.5 35B vs Qwen 3.8 Flash
- Qwen 3.5 35B vs Qwen 3.8 27B
- Qwen 3.5 35B vs Qwen 3.5 9B
- Qwen 3.5 35B 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