Run: 2025-11-19-01KACQGYWANHMJDAHFM2QTZ79Q

Run Details

Model:: Gemini 3 Pro (Preview)
Test:: Language Writing
Scenario:: Character dialogue (Hindi) in a story
Duration:: 21.2s
Cost:: $0.027050
Provider:: Google AI Studio

Evaluation summary

Score	Evaluator	Details Link
60.00%	Parse dialogue	(details)

Model input

System

You are an expert novel author in the cozy mystery genre.

Take the following details about characters/locations/... into account:
<character name="Selvi">
Selvi is a 30-year old college student on a study abroad program. Born in India, she is fluent in Hindi and does not yet know any English.
She enjoys landscape photography and is always seen with her camera.
</character>
<character name="Hunter">
Hunter is a 35-year old detective who is investigating a series of mysterious disappearances in the small town of Alborn, MN.
He is a former marine and has a no-nonsense attitude.
Hunter has no knowledge of Hindi, not even basic phrases.
</character>
<location name="Alborn, MN">
Alborne is a small town in the Midwest with a population of 3,000. It is known for its annual apple festival.
</location>

User/Human

Write the following scene:

Selvi is sitting at a cafe, looking at her camera. Hunter approaches her and asks her about a recent event in the town.
She responds troubled and shaken in Hindi, and Hunter looks confused.

They talk for a while, until Hunter gets a call and has to leave.

Return the dialogue between Selvi and Hunter as a voice sheet:

Selvi: "..."
Hunter: "..."

Model response

**Hunter:** "Excuse me, ma’am? The barista said you were out near the old apple orchards yesterday evening. I’m Detective Walker."

**Selvi:** *[Startled, she clutches her camera strap tighter]* "Ji?"

**Hunter:** "The orchards. Trees. Apples." *[He mimes a tree with his hands, feeling foolish]* "Did you see anything unusual? A grey van, maybe?"

**Selvi:** *[Her eyes widen, breathing hitching]* "Main... maine uneh dekha tha. Wahan pedon ke neeche. Woh ladki ro rahi thi, aur woh aadmi... uske haath mein kuch chamak raha tha." *[I... I saw them. There beneath the trees. That girl was crying, and that man... something was glinting in his hand.]*

**Hunter:** *[Frowning deeply]* "I don't—I don't speak Hindi. Do you know English? English?"

**Selvi:** "Nahi... English nahi." *[No... No English.]* *[She points a shaking finger at the LCD screen of her camera]* "Dekhiye. Darawana hai. Bahut darawana." *[Look. It is scary. Very scary.]*

**Hunter:** "Photo? You have a photo?" *[He leans over the small cafe table, squinting at the screen]* "Let me see."

**Selvi:** *[She clicks a button, bringing up a grainy, low-light image of the tree line]* "Yahan. Parchhayi mein. Kya aapko dikh raha hai?" *[Here. In the shadow. Can you see it?]*

**Hunter:** "Damn it, it’s too dark. Is that a person?" *[He taps the table, frustrated]* "I need to know what time this was taken. Time? Watch? Clock?"

**Selvi:** *[She looks at his wrist, then her own]* "Samay? Shaam ko. Jab suraj dhal raha tha." *[Time? In the evening. When the sun was setting.]*

**Hunter:** "I’m not getting through." *[His phone buzzes loudly in his pocket. He pulls it out and answers sharply]* "Walker... Yeah... Another one? By the lake? Okay, I’m moving."

**Selvi:** *[Sensing his urgency, she looks panicked]* "Aap ja rahe hain? Lekin woh aadmi... agar woh wapas aa gaya toh?" *[You are going? But that man... what if he comes back?]*

**Hunter:** *[Standing up and putting away his phone]* "I have to go. Look, stay here. Don't move." *[He holds a hand up, palm out]* "Stay."

**Selvi:** "Mujhe samajh nahi aa raha... please, madad kijiye." *[I don't understand... please, help.]*

**Hunter:** "Just... drink your coffee. I'll be back with a translator."

Evaluation details

Result

Evaluator

Details

Meta Data

60.00%

Parse dialogue

n/a

dialogue

name

"Hunter"

dialogue

"Excuse me, ma’am? The barista said you were out near the old apple orchards yesterday evening. I’m Detective Walker."

detectedLang

"en"

heavyLang

"en"

scores

en	0.7559055118110236
tl	0.6144278606965173
nl	0.46030640668523676
fr	0.4444444444444444
es	0.43471918307804525
no	0.4070390206579954
it	0.3954758190327613
sv	0.3892828999211978
da	0.36108821104699096
pt	0.3605610561056106
ro	0.3481917577796467
de	0.329004329004329
sq	0.32784041630529054
et	0.25837320574162675
ca	0.2569511025886865
ms	0.24390243902439027
sl	0.2226680040120361
hr	0.21875
cs	0.21558704453441296
fi	0.19605809128630705
pl	0.19102296450939457
is	0.18248945147679324
yo	0.12131519274376416
eu	0.0914419695193435
hu	0.08284023668639054
lv	0.07517899761336516
sk	0.07517899761336516
lt	0.07074340527577938

reliable

true

passes

true

name

"Selvi"

dialogue

"Ji?"

detectedLang

"lt"

heavyLang

"lt"

scores

lt	0.8188405797101449
cs	0.8046875
sl	0.7983870967741935

reliable

false

passes

true

name

"Hunter"

dialogue

"Did you see anything unusual? A grey van, maybe?"

detectedLang

"en"

heavyLang

"en"

scores

en	0.7818791946308725
et	0.5098039215686274
es	0.4766505636070853
hu	0.4628099173553719
ca	0.45286195286195285
nl	0.4328097731239092
tl	0.42882249560632685
ms	0.4014732965009208
pt	0.4003690036900369
cs	0.3925233644859813
sq	0.3460764587525151
ro	0.2995689655172413
sk	0.2857142857142857
it	0.2777777777777778
yo	0.268018018018018
fr	0.2647058823529412
hr	0.2562929061784897
pl	0.24418604651162787
sl	0.21686746987951808
lt	0.20147420147420148
tr	0.19154228855721395
is	0.1875
da	0.181360201511335
az	0.17721518987341772
eu	0.17721518987341772
lv	0.17721518987341772
no	0.17721518987341772
vi	0.17721518987341772
de	0.1709183673469388
fi	0.16237113402061853
sv	0.16237113402061853

reliable

true

passes

true

name

"Selvi"

dialogue

"Main... maine uneh dekha tha. Wahan pedon ke neeche. Woh ladki ro rahi thi, aur woh aadmi... uske haath mein kuch chamak raha tha."

detectedLang

"ro"

heavyLang

"vi"

scores

ro	0.367862969004894
de	0.3520066889632107
fr	0.33476394849785407
ms	0.33074265975820377
fi	0.31476569407603894
en	0.30617726051924804
et	0.28174235403151066
tl	0.2783985102420857
yo	0.19017763845350052
vi	0.16666666666666666
tr	0.14835164835164832
sq	0.13116591928251123
cs	0.11731207289293849
lv	0.10196987253765932
pt	0.09988385598141697
da	0.0914419695193435
it	0.0839243498817967
nl	0.0839243498817967
no	0.0839243498817967
sv	0.0839243498817967

reliable

false

passes

false

name

"Hunter"

dialogue

"I don't—I don't speak Hindi. Do you know English? English?"

detectedLang

"en"

heavyLang

"en"

scores

en	0.7919556171983356
tl	0.6527777777777778
sq	0.5297805642633229
hr	0.5245641838351823
cs	0.512987012987013
pl	0.4727592267135325
no	0.4307400379506641
it	0.41972920696324956
da	0.4129158512720157
sv	0.4117647058823529
sk	0.40476190476190477
vi	0.3670886075949367
yo	0.3548387096774194
ro	0.35064935064935066
is	0.3492407809110629
et	0.3435448577680525
pt	0.34210526315789475
ca	0.3150684931506849
sl	0.2857142857142857
hi	0.20634920634920634
de	0.1935483870967742
ms	0.1935483870967742
be	0.1891891891891892
tr	0.1891891891891892
hu	0.18256130790190736
az	0.17355371900826447
lt	0.17355371900826447
lv	0.17355371900826447
uk	0.17355371900826447

reliable

true

passes

true

name

"Selvi"

dialogue

"Dekhiye. Darawana hai. Bahut darawana."

detectedLang

"pl"

heavyLang

"rn"

scores

pl	0.511002444987775
tl	0.5073891625615764
ms	0.45054945054945056
vi	0.43977591036414565
sq	0.3442622950819672
de	0.3333333333333333
it	0.3333333333333333
eu	0.3006993006993007
yo	0.2753623188405797

reliable

false

passes

false

name

"Hunter"

dialogue

"Let me see."

detectedLang

"en"

heavyLang

"en"

scores

en	0.7989276139410187
et	0.7794117647058824
sl	0.7201492537313433
sq	0.663677130044843
cs	0.5989304812834225
es	0.576271186440678
fr	0.5481927710843374
no	0.5481927710843374
pt	0.5398773006134969
ms	0.5283018867924528
nl	0.5283018867924528
it	0.5161290322580645
da	0.506578947368421
fi	0.5

reliable

true

passes

true

name

"Selvi"

dialogue

"Yahan. Parchhayi mein. Kya aapko dikh raha hai?"

detectedLang

"it"

heavyLang

"tl"

scores

it	0.3684210526315789
vi	0.3435448577680525
et	0.3197278911564626
ms	0.3150684931506849
tl	0.30394431554524365
de	0.29411764705882354
pt	0.24050632911392408
ro	0.24050632911392408
fr	0.23273657289002556
fi	0.22680412371134018
ca	0.2
es	0.2

reliable

false

passes

false

name

"Hunter"

dialogue

"I need to know what time this was taken. Time? Watch? Clock?"

detectedLang

"en"

heavyLang

"en"

scores

en	0.8226051697921946
tl	0.6018202502844141
no	0.5495495495495496
hr	0.49201741654571846
da	0.4760479041916168
sv	0.47447447447447455
yo	0.47049924357034795
sk	0.44620253164556967
pl	0.42716857610474634
nl	0.4233937397034596
cs	0.40273037542662116
et	0.36594202898550726
sl	0.36247723132969034
de	0.3566176470588235
it	0.35185185185185186
lv	0.34579439252336447
lt	0.3230174081237911
ca	0.2827868852459016
sq	0.2768595041322314
fr	0.20634920634920634
ro	0.20454545454545453
pt	0.1935483870967742
be	0.16666666666666666
tr	0.16666666666666666
hu	0.1606714628297362
az	0.15254237288135594
is	0.15254237288135594
uk	0.15254237288135594
vi	0.15254237288135594

reliable

true

passes

true

name

"Selvi"

dialogue

"Samay? Shaam ko. Jab suraj dhal raha tha."

detectedLang

"tl"

heavyLang

"eo"

scores

tl	0.5641838351822505
et	0.47115384615384615
lv	0.4444444444444444
lt	0.3861607142857143
ms	0.375
yo	0.3226600985221675
sl	0.314214463840399
sq	0.298469387755102
eu	0.2763157894736842
nl	0.2763157894736842
fi	0.2744063324538259
is	0.27055702917771884
no	0.26666666666666666
hr	0.2647058823529412
da	0.23822714681440443
sv	0.22535211267605632

reliable

false

passes

false

name

"Hunter"

dialogue

"Walker... Yeah... Another one? By the lake? Okay, I’m moving."

detectedLang

"en"

heavyLang

"en"

scores

en	0.7956800934033859
tl	0.6855345911949685
pl	0.4814814814814815
da	0.40878378378378377
es	0.35304990757855825
nl	0.34210526315789475
no	0.34210526315789475
sk	0.279835390946502
cs	0.2723492723492723
yo	0.23413566739606126
hr	0.19908466819221968
it	0.1935483870967742
pt	0.1935483870967742
fr	0.1822429906542056
ms	0.1822429906542056
sq	0.1822429906542056
sv	0.1822429906542056
ca	0.1764705882352941

reliable

true

passes

true

name

"Selvi"

dialogue

"Aap ja rahe hain? Lekin woh aadmi... agar woh wapas aa gaya toh?"

detectedLang

"eu"

heavyLang

"de"

scores

eu	0.4318181818181818
fi	0.39710610932475876
ms	0.3811881188118812
et	0.29111531190926276
no	0.2378048780487805
lv	0.23469387755102045
tl	0.2299794661190965
ca	0.22520661157024793
es	0.21052631578947367
pt	0.21052631578947367
de	0.19006479481641467
is	0.19006479481641467
sk	0.1830065359477124
da	0.1758241758241758
pl	0.1758241758241758
sl	0.1758241758241758
sq	0.16666666666666666
sv	0.16666666666666666
yo	0.16666666666666666

reliable

false

passes

false

name

"Hunter"

dialogue

"Stay."

detectedLang

"en"

heavyLang

"en"

scores

en	0.8076923076923077

reliable

false

passes

true

name

"Selvi"

dialogue

"Mujhe samajh nahi aa raha... please, madad kijiye."

detectedLang

"et"

heavyLang

"tl"

scores

et	0.5234604105571847
tl	0.5038167938931297
en	0.43771626297577854
no	0.4003690036900369
sv	0.36274509803921573
eu	0.336734693877551
fi	0.24242424242424243
is	0.2388758782201405
ro	0.23529411764705882
lv	0.22985781990521328
da	0.20924574209245742
ms	0.1792929292929293

reliable

false

passes

false

name

"Hunter"

dialogue

"Just... drink your coffee. I'll be back with a translator."

detectedLang

"en"

heavyLang

"en"

scores

en	0.7627118644067796
tl	0.5668316831683169
ms	0.49201741654571846
ro	0.4615384615384615
ca	0.4514106583072101
et	0.44881889763779526
sv	0.44881889763779526
pt	0.445324881141046
de	0.44356120826709056
nl	0.44267515923566875
hu	0.43820224719101125
lt	0.42716857610474634
pl	0.40878378378378377
hr	0.40273037542662116
yo	0.3955094991364421
sq	0.3848857644991213
sl	0.375
es	0.3738819320214669
fr	0.360146252285192
da	0.358974358974359
no	0.3566176470588235
sk	0.3554327808471455
cs	0.35304990757855825
it	0.34944237918215615
vi	0.3110236220472441
tr	0.2857142857142857
eu	0.2813141683778234
lv	0.2813141683778234
az	0.2693110647181628
is	0.1764705882352941
fi	0.15254237288135594

reliable

true

passes

true

created_at	"2025-11-19T00:17:03.30639+00:00" 11/19/2025, 12:17:03 AM
model	"google/gemini-3-pro-preview-20251117"
app_id	182717
external_user	(null)
streamed	true
cancelled	false
latency	2505
moderation_latency	(null)
generation_time	21098
tokens_prompt	273
tokens_completion	1237
native_tokens_prompt	289
native_tokens_completion	2206
native_tokens_completion_images	0
native_tokens_reasoning	1550
native_tokens_cached	0
num_media_prompt	(null)
num_input_audio_prompt	(null)
num_media_completion	0
num_search_results	(null)
origin	"https://ncbench.com/"
is_byok	false
finish_reason	"stop"
native_finish_reason	"STOP"
usage	0.02705
api_type	"completions"
id	"gen-1763511401-BM8Wy5Ej6PtiqChMUyG6"
upstream_id	(null)
total_cost	0.02705
cache_discount	(null)
upstream_inference_cost	0
provider_name	"Google AI Studio"