Skip to content

Model qwen3 14b

hydropix edited this page Dec 23, 2025 · 1 revision

qwen3:14b

Ollama Model ID: qwen3:14b


Summary

Metric Value
Average Score 🟠 6.0/10
Accuracy 6.7/10
Fluency 6.0/10
Style 5.7/10
Languages Tested 19
Total Translations 95
Best Language Portuguese (7.4)
Worst Language Hebrew (1.8)

Language Performance

Top Languages

Rank Language Overall Accuracy Fluency Style
1 Portuguese 🟡 7.4 8.2 7.4 7.2
2 French 🟡 7.2 7.8 7.2 6.8
3 Italian 🟡 7.2 7.4 7.2 6.6
4 Chinese (Traditional) 🟡 7.2 7.6 7.2 7.0
5 Chinese (Simplified) 🟡 7.0 7.6 7.2 6.8
6 Spanish 🟠 6.8 7.2 7.0 6.4
7 German 🟠 6.4 7.2 6.2 6.2
8 Russian 🟠 6.4 7.0 6.6 6.2
9 Polish 🟠 6.2 6.8 6.0 5.8
10 Arabic 🟠 6.2 6.8 6.2 5.8
View all 19 languages
Rank Language Overall Accuracy Fluency Style
1 Portuguese 🟡 7.4 8.2 7.4 7.2
2 French 🟡 7.2 7.8 7.2 6.8
3 Italian 🟡 7.2 7.4 7.2 6.6
4 Chinese (Traditional) 🟡 7.2 7.6 7.2 7.0
5 Chinese (Simplified) 🟡 7.0 7.6 7.2 6.8
6 Spanish 🟠 6.8 7.2 7.0 6.4
7 German 🟠 6.4 7.2 6.2 6.2
8 Russian 🟠 6.4 7.0 6.6 6.2
9 Polish 🟠 6.2 6.8 6.0 5.8
10 Arabic 🟠 6.2 6.8 6.2 5.8
11 Vietnamese 🟠 6.0 7.0 6.0 6.0
12 Thai 🟠 6.0 7.0 6.0 6.0
13 Ukrainian 🟠 6.0 6.8 6.0 5.8
14 Korean 🟠 5.6 6.4 5.8 5.2
15 Japanese 🟠 5.4 6.2 5.4 4.8
16 Hindi 🟠 5.2 6.2 5.2 4.8
17 Bengali 🔴 4.8 5.8 4.6 4.6
18 Tamil 🔴 4.8 5.8 4.8 4.2
19 Hebrew ⚫ 1.8 2.8 2.0 1.8

Performance by Category

European Major Languages

Language Overall Accuracy Fluency Style
Portuguese 🟡 7.4 8.2 7.4 7.2
French 🟡 7.2 7.8 7.2 6.8
Italian 🟡 7.2 7.4 7.2 6.6
Spanish 🟠 6.8 7.2 7.0 6.4
German 🟠 6.4 7.2 6.2 6.2
Polish 🟠 6.2 6.8 6.0 5.8

Category Average: 🟠 6.9

Asian Languages

Language Overall Accuracy Fluency Style
Chinese (Traditional) 🟡 7.2 7.6 7.2 7.0
Chinese (Simplified) 🟡 7.0 7.6 7.2 6.8
Vietnamese 🟠 6.0 7.0 6.0 6.0
Thai 🟠 6.0 7.0 6.0 6.0
Korean 🟠 5.6 6.4 5.8 5.2
Japanese 🟠 5.4 6.2 5.4 4.8
Hindi 🟠 5.2 6.2 5.2 4.8
Bengali 🔴 4.8 5.8 4.6 4.6
Tamil 🔴 4.8 5.8 4.8 4.2

Category Average: 🟠 5.8

Cyrillic Languages

Language Overall Accuracy Fluency Style
Russian 🟠 6.4 7.0 6.6 6.2
Ukrainian 🟠 6.0 6.8 6.0 5.8

Category Average: 🟠 6.2

Semitic Languages

Language Overall Accuracy Fluency Style
Arabic 🟠 6.2 6.8 6.2 5.8
Hebrew ⚫ 1.8 2.8 2.0 1.8

Category Average: 🔴 4.0


Translation Examples

Best Translation

French - Walden Score: 🟡 8.0/10

View Translation

Original:

Translation:

Je suis allé dans les bois parce que je voulais vivre délibérément, ne faire face qu'aux faits essentiels de la vie, et voir si je ne pourrais pas apprendre ce qu'elle avait à m'apprendre, et ne pas, lors de ma mort, découvrir que je n'avais pas vraiment vécu. Je ne voulais pas vivre ce qui n'était pas la vie, la vie étant si précieuse ; ni pratiquer la résignation, sauf si cela devenait absolument nécessaire. Je voulais vivre profondément et tirer toutes les moelles de la vie.

Feedback: The translation faithfully preserves Thoreau's meaning and philosophical intent with only minor deviations (e.g., 'vraiment' adds emphasis not in original, 'tirer toutes les moelles' is slightly awkward compared to 'suck out all the marrow'). The prose flows naturally in French, though some of the original's rhythmic intensity is slightly diminished in the final clause.

Worst Translation

Hebrew - Pride and Prejudice Score: ⚫ 1.0/10

View Translation

Original:

Translation:

האמת המוכחת באופן כללי היא שرجل נשוי שבעלתו מוניטין טוב חייב להיות בצרה של נישום. עם זאת, גם אם הרגשות והעמדות של איש כזה ידועות מעט מאוד כאשר הוא פונה לראשונה ל الحي, האמת הזו נשמרת היטב בחובשות של המשפחות הסובבות אותו, כך שהוא נחשב לממון מוצדק של אחת מה딸ות שלהם.

Feedback: This translation is severely compromised by multiple critical errors: it contains untranslated Arabic and Korean text mixed into Hebrew, fundamentally misrepresents the meaning (changing 'single man' to 'married man'), uses incorrect Hebrew vocabulary ('נישום' is not a real word), and completely fails to preserve Austen's ironic tone and elegant prose structure. The translation is essentially unusable.


Score Distribution

Score Range Count Percentage
🟢 Excellent (9-10) 0 0.0%
🟡 Good (7-8) 39 41.1%
🟠 Acceptable (5-6) 45 47.4%
🔴 Poor (3-4) 7 7.4%
⚫ Failed (1-2) 4 4.2%

Performance Metrics

Metric Value
Average Translation Time 10344.0ms
Success Rate 100.0%

← Back to Home | All Models

Clone this wiki locally