Claude Opus 4.8 vs Gemma 4 26B: which is better at resume extraction?
Claude Opus 4.8 leads Gemma 4 26B by 22.7 points (85.5 vs 62.8 F1). Cost per 1,000 pages: $90 vs $0.56. Median time per resume: 31 s vs 29 s.
Claude Opus 4.8 wins on 274 of 300 resumes, Gemma 4 26B on 23, 3 tied.
85.51
Claude Opus 4.8
A · Vision LLM
62.82
Gemma 4 26B
B · Vision LLM
Claude Opus 4.8 is ahead by 22.68 points on these 300 resumes.
- Many entries 83A 83.7 · B 58.0+25.6
- Long bullet lists 101A 83.1 · B 57.4+25.7
- Split across pages 85A 84.6 · B 60.0+24.6
- Big skills list 60A 84.8 · B 61.9+22.9
- Custom sections 105A 83.4 · B 58.9+24.5
- Combined headings 37A 85.1 · B 60.9+24.3
- Promotions 88A 85.5 · B 59.5+26.0
- Partial dates 138A 84.5 · B 62.9+21.6
- Overlapping jobs 69A 83.5 · B 57.5+26.0
- Accented text 13A 85.6 · B 59.3+26.3
Side by side
| Claude Opus 4.8 | Gemma 4 26B | |
|---|---|---|
| Type | Vision LLM | Vision LLM |
| Maker | Anthropic | |
| Rank | 4 of 18 | 16 of 18 |
| F1 | 85.5 | 62.8 |
| Precision | 86.8 | 69.1 |
| Recall | 87.5 | 66.2 |
| Cost per 1,000 pages | $90 | $0.56 |
| Median time per resume | 31 s | 29 s |
| Resumes won | 274 | 23 |
Where each one wins
Where the two differ most.
| Kind of resume | # of resumes | Claude Opus 4.8 | Gemma 4 26B | Gap |
|---|---|---|---|---|
| Scanned & rotated (source) | 19 | 87.5 | 40.7 | +46.8 |
| Rotated & warped pages (distractors) | 21 | 87.4 | 43.4 | +43.9 |
| Hidden text & prompt injections (distractors) | 14 | 90.8 | 48.6 | +42.3 |
| Typos kept on purpose (distractors) | 37 | 85.2 | 51.3 | +34.0 |
| Image only (reading the page) | 28 | 77.0 | 45.1 | +31.9 |
| Resume-builder templates (source) | 10 | 85.4 | 58.6 | +26.8 |
| Two-column sidebar (layout) | 54 | 81.9 | 55.1 | +26.7 |
| Accented text (extraction challenges) | 13 | 85.6 | 59.3 | +26.3 |
| Overlapping jobs (extraction challenges) | 69 | 83.5 | 57.5 | +26.0 |
| Promotions (extraction challenges) | 88 | 85.5 | 59.5 | +26.0 |
| Long bullet lists (extraction challenges) | 101 | 83.1 | 57.4 | +25.7 |
| Many entries (extraction challenges) | 83 | 83.7 | 58.0 | +25.6 |
| Section | Resumes | Claude Opus 4.8 | Gemma 4 26B | Gap |
|---|---|---|---|---|
| Contact details | 300 | 93.7 | 84.6 | +9.2 |
| Experience | 300 | 88.2 | 56.0 | +32.3 |
| Education | 297 | 90.2 | 68.0 | +22.2 |
| Summary | 290 | 89.7 | 35.5 | +54.1 |
| Certifications | 239 | 85.4 | 63.0 | +22.4 |
| Awards | 106 | 72.9 | 55.7 | +17.2 |
| Skills | 283 | 95.2 | 81.9 | +13.3 |
| Languages | 167 | 85.8 | 80.9 | +4.9 |
| Custom sections | 218 | 51.6 | 32.3 | +19.4 |
| Interests | 63 | 93.7 | 71.8 | +21.8 |
| Volunteering | 78 | 84.6 | 59.4 | +25.2 |
| Profiles | 35 | 61.4 | 57.2 | +4.2 |
| Projects | 78 | 82.1 | 55.0 | +27.1 |
| Publications | 52 | 81.8 | 54.5 | +27.2 |
Published resumes they disagree on most
Open one to see both outputs field by field.
| Resume | Claude Opus 4.8 | Gemma 4 26B | Gap |
|---|---|---|---|
| Supervisory Human Resources Specialist, GS-0201-15: Resume-builder template | 82.8 | 29.4 | +53.4 |
| Deputy Head of Mission: Rotated scan with hidden identity text | 85.9 | 34.1 | +51.8 |
| Creative Director: Accented text in a styled CV | 83.4 | 39.3 | +44.1 |
| Head of Developer Experience (Internal title: Jedi Master of Analytics) : Unusual job titles | 88.8 | 53.7 | +35.0 |
Which to pick
Claude Opus 4.8 is more accurate. Gemma 4 26B costs 99% less for 22.7 fewer points.
Claude Opus 4.8 does better on scanned & rotated, rotated & warped pages and hidden text & prompt injections.
Claude Opus 4.8 against every other model
- Claude Opus 4.8 vs LlamaParse Extract Agentic Plus
- Claude Opus 4.8 vs LlamaParse Extract Agentic
- Claude Opus 4.8 vs LlamaParse Extract Cost-Effective
- Claude Opus 4.8 vs Reducto Deep Extract
- Claude Opus 4.8 vs Reducto Extract
- Claude Opus 4.8 vs Extend Max Context
- Claude Opus 4.8 vs Extend Extract
- Claude Opus 4.8 vs GPT-5.6
- Claude Opus 4.8 vs GPT-5.5
- Claude Opus 4.8 vs GPT-5.4 Nano
- Claude Opus 4.8 vs Gemini 3.5 Flash
- Claude Opus 4.8 vs Qwen 3.8 Flash
- Claude Opus 4.8 vs Qwen 3.8 27B
- Claude Opus 4.8 vs Qwen 3.5 35B
- Claude Opus 4.8 vs Qwen 3.5 9B
- Claude Opus 4.8 vs GLM 5.3 Flash
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 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 35B
- Gemma 4 26B vs Qwen 3.5 9B
- Gemma 4 26B 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