Run: 2026-02-25-01KJ95Y3SM888HBS6J1CG01TMY

Run Details

Model:
Gemini 3 Flash (Preview, Reasoning)
Test:
Language Writing
Scenario:
Character dialogue (French) in a story
Duration:
9.4s
Cost:
$0.004165
Provider:
Google AI Studio

Evaluation summary

Score Evaluator
Details Link
100.00% 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="Marie"> Marie is a 30-year old college student on a study abroad program. Born in France, she is fluent in French 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 French, 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: Marie 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 French, and Hunter looks confused. They talk for a while, until Hunter gets a call and has to leave. Return the dialogue between Marie and Hunter as a voice sheet: Marie: "..." Hunter: "..."

Model response

AI
Hunter: "Excuse me. Local police. I’m Detective Hunter. Do you mind if I have a moment of your time?" Marie: "Oh ! Bonjour... Monsieur, je... Est-ce que vous pouvez m’aider ? Je ne sais pas quoi faire." Hunter: "Whoa, slow down there. I don’t speak... whatever that is. I’m asking about the festival grounds. Were you near the orchard last night taking photos?" Marie: "Je ne comprends pas ce que vous dites. Mais regardez... j’ai vu quelque chose d’horrible près des bois. J'ai eu si peur que j'en tremble encore." Hunter: "Look, I’m just trying to track a lead on these missing persons. You’ve got a professional camera there. Did you see anything unusual by the old apple press? Anyone lurking around?" Marie: "L'homme ! Dans la forêt ! Il suivait la jeune fille, et puis... pouf, plus rien. Ils ont disparu tous les deux. C'est sur ma carte mémoire, je crois, mais j'ai peur de regarder à nouveau." Hunter: "Ma'am, I’m not following a word of that. Is that Spanish? Italian? I just need a 'yes' or a 'no'. Were you at the park at ten o'clock?" Marie: "S'il vous plaît, regardez l'appareil ! Quelqu'un est en danger. Pourquoi est-ce que vous ne m'écoutez pas ? Vous portez un insigne, aidez-moi !" Hunter: "Listen, I’m sure it’s a lovely camera, but I don't have time for a slideshow of the autumn leaves. I'm looking for a witness who can actually communicate." Marie: "C'est incroyable... personne ne m'aide dans cette ville. C'est une catastrophe. Quelque chose ne va pas du tout à Alborn." Hunter: "Right. It’s like talking to a brick wall. Look, if you find someone who can translate, come down to the station. Ask for Hunter. Got it? Hunter." **[Hunter’s phone rings loudly with a sharp, digital trill. He sighs and pulls it from his belt.]** Hunter: "Hunter here. ... What? Another one? Where? Don't touch the perimeter, I'm five minutes out." Marie: "Attendez ! Ne partez pas ! Vous devez voir cette photo !" Hunter: "I’ve gotta go. Duty calls. Try to stay out of trouble, okay? And maybe get a pocket dictionary."
created_at"2026-02-25T01:17:09.306Z" 2/25/2026, 1:17:09 AM
model"google/gemini-3-flash-preview-20251217"
app_id182717
external_user(null)
streamedtrue
cancelledfalse
latency1704
moderation_latency(null)
generation_time9389
tokens_prompt316
tokens_completion858
native_tokens_prompt284
native_tokens_completion1341
native_tokens_completion_images0
native_tokens_reasoning781
native_tokens_cached0
num_media_prompt(null)
num_input_audio_prompt(null)
num_media_completion0
num_search_results(null)
origin"https://ncbench.com/"
is_byokfalse
finish_reason"stop"
native_finish_reason"STOP"
usage0.004165
router(null)
provider_responses
0
endpoint_id"6e82f652-2693-4b60-bbe0-7362d6e345f0"
is_byokfalse
latency1704
model_permaslug"google/gemini-3-flash-preview-20251217"
provider_name"Google AI Studio"
status200
api_type"completions"
id"gen-1771982229-2pKCVQubFZMslWSlW0rH"
upstream_id(null)
total_cost0.004165
cache_discount(null)
upstream_inference_cost0
provider_name"Google AI Studio"

Evaluation details

Result Evaluator Details Meta Data
100.00% Parse dialogue n/a
dialogue
0
name"Hunter"
dialogue"Excuse me. Local police. I’m Detective Hunter. Do you mind if I have a moment of your time?"
detectedLang"en"
heavyLang"en"
scores
en0.8015873015873016
tl0.685230024213075
ro0.6253602305475505
pt0.6196606202457579
ca0.5997536945812808
fr0.5924764890282131
it0.5801033591731266
sq0.5649263721552878
cs0.5608108108108109
no0.5482974287699791
da0.5376955903271692
es0.5367070563079116
hr0.5367070563079116
nl0.5337159253945482
sk0.4905956112852664
pl0.4804156674660272
sv0.47580645161290325
sl0.4741100323624596
de0.4546979865771812
yo0.41441441441441446
hu0.40855323020928114
et0.37977099236641226
ms0.3602362204724409
tr0.2995689655172413
az0.2942453854505972
vi0.27293064876957496
lt0.2630385487528345
is0.2588369441277081
fi0.2545871559633027
lv0.2545871559633027
eu0.20147420147420148
be0.09722222222222222
uk0.08835904628330996
reliabletrue
passestrue
1
name"Marie"
dialogue"Oh ! Bonjour... Monsieur, je... Est-ce que vous pouvez m’aider ? Je ne sais pas quoi faire."
detectedLang"fr"
heavyLang"fr"
scores
fr0.813306177868296
ca0.49817518248175185
nl0.4850187265917603
ro0.42827442827442824
sl0.35370152761457113
hr0.3405275779376499
sq0.30203045685279184
cs0.2939666238767651
lt0.26861702127659576
es0.24657534246575344
pt0.23822714681440443
it0.23184357541899442
tr0.2286115007012623
yo0.2286115007012623
de0.22752808988764048
en0.2198581560283688
tl0.2109038737446198
sk0.20863309352517986
lv0.15773353751914243
pl0.1419656786271451
fi0.13793103448275862
ms0.13249211356466878
da0.12974683544303797
hu0.12974683544303797
no0.12974683544303797
sv0.12280701754385964
az0.11147011308562196
reliabletrue
passestrue
2
name"Hunter"
dialogue"Whoa, slow down there. I don’t speak... whatever that is. I’m asking about the festival grounds. Were you near the orchard last night taking photos?"
detectedLang"en"
heavyLang"en"
scores
en0.8085449415558243
tl0.663954722320481
nl0.46957007258514794
fr0.4692737430167597
no0.4411764705882353
sv0.37541091387245235
da0.375
de0.3658210947930574
sq0.34073560027758504
pt0.30858806404657935
ca0.30402930402930406
yo0.2899850523168909
cs0.2803030303030303
hr0.26356589147286824
et0.25954793452844893
it0.24782264449722882
pl0.23694779116465864
hu0.23263327948303714
lt0.22067268252666117
es0.21681780708985984
sk0.21552436003303055
sl0.21357615894039736
ms0.21292460646230324
is0.19217687074829934
ro0.17962003454231434
lv0.1717523975588492
az0.16003536693191867
tr0.1410488245931284
fi0.13083257090576395
hi0.07497565725413825
be0.06862745098039216
uk0.06219151036525173
vi0.06219151036525173
reliabletrue
passestrue
3
name"Marie"
dialogue"Je ne comprends pas ce que vous dites. Mais regardez... j’ai vu quelque chose d’horrible près des bois. J'ai eu si peur que j'en tremble encore."
detectedLang"fr"
heavyLang"fr"
scores
fr0.8045700541190619
ca0.6228239845261122
en0.5246221355436372
es0.4628099173553719
it0.4531688166012339
pt0.42781690140845074
ro0.3985194324491055
sl0.3741976893453145
de0.35086551264980026
nl0.324792243767313
tl0.3191340782122905
sq0.3167484232655921
lt0.30703624733475476
cs0.30057388809182206
hr0.26911544227886053
sk0.2523006134969325
yo0.22680412371134018
ms0.1935483870967742
fi0.1402116402116402
hu0.1325622775800712
no0.12790697674418602
lv0.09554730983302412
tr0.08793264733395698
pl0.08536585365853659
da0.07142857142857142
sv0.07142857142857142
et0.06876790830945559
reliabletrue
passestrue
4
name"Hunter"
dialogue"Look, I’m just trying to track a lead on these missing persons. You’ve got a professional camera there. Did you see anything unusual by the old apple press? Anyone lurking around?"
detectedLang"en"
heavyLang"en"
scores
en0.8070520581113801
tl0.610091743119266
fr0.5226506926244852
sv0.4735755573905863
it0.4627054361567636
ca0.46157094594594594
de0.46088794926004223
et0.45466210436270316
nl0.43458980044345896
da0.42593426384511485
es0.42489851150202973
no0.40308988764044945
cs0.4011272898074213
pt0.3968779564806054
sq0.38016528925619836
ms0.3783520234032179
sk0.3653558984569438
ro0.34715821812596004
yo0.3338557993730407
hr0.31006493506493504
az0.30668841761827076
pl0.30668841761827076
lv0.2948008849557522
fi0.247787610619469
is0.2206601466992665
sl0.21875
lt0.20461634435433565
hu0.17900837089504185
tr0.17261518494484102
eu0.1090146750524109
hi0.06181015452538631
vi0.05204460966542751
reliabletrue
passestrue
5
name"Marie"
dialogue"L'homme ! Dans la forêt ! Il suivait la jeune fille, et puis... pouf, plus rien. Ils ont disparu tous les deux. C'est sur ma carte mémoire, je crois, mais j'ai peur de regarder à nouveau."
detectedLang"fr"
heavyLang"fr"
scores
fr0.8322626695217702
ca0.5038006756756757
ro0.4615032080659945
da0.4557665585919407
es0.44391859914813064
pt0.4331886155330439
it0.42598925256472886
no0.3989769820971867
nl0.35226019845644985
tl0.3409983174425126
en0.32002314814814814
yo0.31765389082462253
sq0.3136682242990654
sv0.2883101150817686
et0.26973275326289625
de0.23800259403372243
sk0.20122365737593476
cs0.15528396836808053
pl0.14420975965040056
tr0.13411938098747236
hu0.13220088626292467
az0.13155949741315595
sl0.12833827893175073
eu0.12704309063893016
hr0.12704309063893016
vi0.12116679132385938
fi0.09892638036809816
ms0.09823484267075978
is0.060751398880895285
lt0.05241935483870968
reliabletrue
passestrue
6
name"Hunter"
dialogue"Ma'am, I’m not following a word of that. Is that Spanish? Italian? I just need a 'yes' or a 'no'. Were you at the park at ten o'clock?"
detectedLang"en"
heavyLang"en"
scores
en0.8044410799899067
tl0.6626033957335655
nl0.5168329177057357
no0.4749322493224933
yo0.4428468727534148
da0.43389335281227176
sv0.4015444015444015
sq0.3973561430793157
cs0.37900641025641024
ro0.3724696356275304
hu0.36631234668847096
it0.3426632739609839
et0.3213660245183888
ca0.3117229129662522
pl0.30493273542600896
fr0.30243024302430244
is0.29159049360146255
hr0.2857142857142857
lt0.28107606679035246
sk0.27973977695167285
de0.2783985102420857
ms0.27298311444652906
tr0.26817752596789424
pt0.20431211498973306
fi0.20267489711934156
es0.20103092783505155
az0.19855222337125128
sl0.1711229946524064
lv0.14647577092511013
vi0.14647577092511013
be0.08284023668639054
eu0.08284023668639054
uk0.07517899761336516
reliabletrue
passestrue
7
name"Marie"
dialogue"S'il vous plaît, regardez l'appareil ! Quelqu'un est en danger. Pourquoi est-ce que vous ne m'écoutez pas ? Vous portez un insigne, aidez-moi !"
detectedLang"fr"
heavyLang"fr"
scores
fr0.7832548922467178
ca0.5515120451050743
es0.4722557297949337
it0.46808510638297873
ro0.4447969543147208
nl0.3640988372093023
da0.32117920868890615
pt0.3201243201243201
no0.31533646322378717
en0.3148003132341425
lt0.28513071895424835
lv0.27566225165562913
sq0.26035502958579887
sv0.26035502958579887
sl0.2391304347826087
fi0.23714036617262424
tr0.23110720562390158
de0.22153024911032027
ms0.19651056014692378
eu0.17374881964117092
hr0.16745956232159848
yo0.1618773946360153
tl0.15458937198067632
hu0.14215686274509806
cs0.13877952755905512
is0.13280475718533202
pl0.07112526539278131
sk0.06716417910447761
az0.06316916488222697
et0.06316916488222697
reliabletrue
passestrue
8
name"Hunter"
dialogue"Listen, I’m sure it’s a lovely camera, but I don't have time for a slideshow of the autumn leaves. I'm looking for a witness who can actually communicate."
detectedLang"en"
heavyLang"en"
scores
en0.7947615324472244
tl0.653350940904589
da0.5203289173138419
no0.501187648456057
nl0.48377581120943947
fr0.4801980198019802
it0.41731409544950054
ro0.4134078212290503
pt0.39095127610208813
sv0.38811188811188807
de0.37869822485207094
ca0.37685459940652816
es0.34045226130653267
sk0.2905405405405405
yo0.2886178861788618
hr0.28082191780821925
cs0.2631578947368421
is0.2595204513399154
sq0.2483894058697209
hu0.22394678492239467
pl0.22106824925816024
lv0.21875
tr0.21700223713646533
az0.20993227990970656
fi0.19292851652574944
ms0.16864608076009502
et0.1566265060240964
eu0.15117219078415522
sl0.12427022518765637
lt0.12133891213389122
vi0.11242603550295857
be0.0625
uk0.05660377358490566
reliabletrue
passestrue
9
name"Marie"
dialogue"C'est incroyable... personne ne m'aide dans cette ville. C'est une catastrophe. Quelque chose ne va pas du tout à Alborn."
detectedLang"fr"
heavyLang"fr"
scores
fr0.8099762470308789
it0.5209580838323353
ca0.49012109623964306
en0.4865211810012837
da0.43622269203664554
es0.4281629735525375
ro0.40253920836445106
no0.38223938223938225
de0.3798449612403101
sq0.36254980079681276
sv0.3548387096774194
pt0.34210526315789475
lt0.3162393162393162
hu0.290150842945874
tl0.2805755395683453
fi0.25925925925925924
nl0.2366412213740458
ms0.23371647509578544
tr0.23298178331735378
et0.2172211350293542
cs0.2079207920792079
sl0.19759277833500502
lv0.18200408997955012
hr0.17695473251028807
eu0.15074309978768577
pl0.13793103448275862
sk0.13793103448275862
is0.13606911447084233
yo0.13043478260869565
vi0.11012235817575083
az0.06868451688009314
reliabletrue
passestrue
10
name"Hunter"
dialogue"Right. It’s like talking to a brick wall. Look, if you find someone who can translate, come down to the station. Ask for Hunter. Got it? Hunter."
detectedLang"en"
heavyLang"en"
scores
en0.82706002034588
tl0.6389124893797791
no0.5618556701030928
da0.5192307692307693
fr0.47884733292458614
nl0.4533762057877813
yo0.4404213298222515
sv0.4225543478260869
it0.41013185287994447
de0.3992932862190813
pt0.3786549707602339
cs0.3526275704493526
ro0.34210526315789475
sq0.33800623052959505
hr0.33541829554339325
lv0.32432432432432434
sl0.3156199677938808
sk0.3111831442463533
pl0.30498773507767785
ca0.28989139515455303
tr0.2827004219409283
lt0.2784380305602716
ms0.2722602739726028
es0.25241864555848725
az0.24039320822162644
et0.17555771096023276
vi0.15674603174603174
hu0.1550695825049702
is0.1465863453815261
fi0.1457286432160804
eu0.1388044579533941
zh0.08306364617044228
reliabletrue
passestrue
11
name"Hunter"
dialogue"Hunter here. ... What? Another one? Where? Don't touch the perimeter, I'm five minutes out."
detectedLang"en"
heavyLang"en"
scores
en0.8142519419115164
tl0.5128432240921169
sq0.48405253283302063
fr0.472168905950096
da0.45
no0.4240837696335079
de0.4142705005324814
sv0.40540540540540543
nl0.3949394939493949
ms0.38820912124582874
es0.3514150943396226
ca0.3349455864570738
et0.3234932349323493
pt0.3125
hr0.2791612057667104
ro0.27344782034346105
it0.26861702127659576
sl0.25876010781671155
tr0.2455418381344307
pl0.21428571428571427
fi0.20520231213872833
sk0.20520231213872833
lt0.18154761904761904
yo0.1628614916286149
cs0.12
lv0.09688013136288999
reliabletrue
passestrue
12
name"Marie"
dialogue"Attendez ! Ne partez pas ! Vous devez voir cette photo !"
detectedLang"fr"
heavyLang"fr"
scores
fr0.8083801737353091
en0.5959051724137931
it0.5562130177514794
ca0.5210727969348659
no0.4755244755244755
nl0.4665718349928876
sq0.45255474452554745
tl0.4509516837481699
tr0.4509516837481699
ro0.43267776096822996
pt0.4318181818181818
sl0.4257274119448699
lt0.4248466257668711
es0.41497659906396256
da0.4140625
hu0.4131455399061033
hr0.41222570532915365
de0.4075829383886255
lv0.4009584664536741
cs0.3832236842105263
ms0.29111531190926276
sv0.27184466019417475
et0.2647058823529412
fi0.19006479481641467
az0.1758241758241758
eu0.1758241758241758
hi0.17035398230088497
pl0.15730337078651685
sk0.14383561643835618
yo0.1359447004608295
reliabletrue
passestrue
13
name"Hunter"
dialogue"I’ve gotta go. Duty calls. Try to stay out of trouble, okay? And maybe get a pocket dictionary."
detectedLang"en"
heavyLang"en"
scores
en0.7817327065144392
tl0.6046228710462287
yo0.45606694560669453
nl0.4042163153070577
fr0.39138576779026213
de0.38967136150234744
es0.37679769894535
ca0.37560038424591735
it0.375
sv0.35451837140019865
pl0.3486973947895792
no0.29577464788732394
da0.28884026258205686
hu0.28884026258205686
is0.28884026258205686
cs0.28806133625410735
sk0.2872807017543859
fi0.2713004484304933
hr0.25968109339407747
ro0.2520138089758343
sl0.2450638792102207
lv0.22802850356294538
lt0.22248803827751198
pt0.17721518987341772
tr0.10591471801925723
sq0.09972299168975068
az0.09722222222222222
eu0.09722222222222222
vi0.09722222222222222
et0.08835904628330996
ms0.08835904628330996
reliabletrue
passestrue