Run: 2026-02-16-01KHKATNPKQBKV0ZZQH45HX65R

Run Details

Model:
GPT-5 Mini
Test:
Language Writing
Scenario:
Character dialogue (Hindi) in a story
Duration:
38.6s
Cost:
$0.007614
Provider:
OpenAI

Evaluation summary

Score Evaluator
Details Link
55.00% 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

AI
Hunter: "(approaching, professional) Hi—are you Selvi? I'm Detective Hunter. Mind if I ask you a few questions about the recent disappearances in town?" Selvi: "(hands trembling on her camera, in Hindi) Haan. Main Selvi hoon. (long breath) Main theek nahin hoon. Meri camera ki tasveeren... mujhe bahut dar rahi hain." Hunter: "(politely insistent) Okay—can you tell me what you saw? Anything at all. Names, places?" Selvi: "(shaken, in Hindi) Pichli raat main nadi ke kinare photo le rahi thi. Maine ek aadmi dekha—woh bahut pareshaan tha. Maine socha nazar se chalu ho jayega, isliye maine photo li." Hunter: "(leans forward, quietly) Can I—can I see the pictures?" Selvi: "(hurried, in Hindi) Nahi... abhi nahin. Main dar gayi hoon. Photo mein us aadmi ke chehre par kuch tha—jaise nishaan, aur phir subah uska ghar khaali tha." Hunter: "(confused, trying) I'm sorry—I don't understand. Who's missing? A friend? A neighbor?" Selvi: "(urgent, in Hindi) Uska naam—maine poora nahin poocha. Kisi ne use dekhna chhod diya. Ab koi usse dekh nahin raha. Main police ko bataana chahti thi, par main angrezi nahin bol sakti." Hunter: "(takes out notebook, practical) You don't speak English at all?" Selvi: "(quiet, in Hindi) Nahin. Main abhi-abhi yahan aayi hoon. Main Hindi bolti hoon." Hunter: "(softens, trying to bridge) Okay. Point to where you took the photos. Show me on the camera—please." Selvi: "(nods quickly, hands him the camera without speaking, then in Hindi) Yahi—yahi river ke paas. Yeh pictures—maine dekha aur tab se darr raha hoon." Hunter: "(fumbles with the camera, then looks up) I... I can't read captions. Can you point at the person in the photo?" Selvi: "(pointing on the screen, agitated, in Hindi) Wahan! Wahan! Uska haath... woh kisi cheez ko pakad raha tha. Aur uske peeche... jaise koi saaya ho." Hunter: "(increasingly concerned) Shadow? Okay. Is this person someone from town? Do you know him?" Selvi: "(quiet, in Hindi) Main usse pahchaan nahin paayi. Par jab main raat ko wapas gayi to uska ghar khaali tha. Darr lagta hai—main samajh nahin pa rahi." Hunter: "(steady, low) Stay right here. Don't move. I'll get a translator and we'll go through every picture. Do not delete anything." Selvi: "(grasping his sleeve, pleading in Hindi) Kripya—jaldi. Koi wapas aa jaaye... main dar se so nahin pa rahi." Hunter: "(firm but gentle) I will—I'll get help. I'm not leaving you. Just stay put." Hunter: "(phone buzzes; he answers, brisk) Hold on—I've got to take this. (into phone) Hunter. ...Yes. On my way. (to Selvi) I have to go—call came from the other disappearance. I'll be right back with someone who can translate. Don't go anywhere."
created_at"2026-02-16T13:40:01.260032+00:00" 2/16/2026, 1:40:01 PM
model"openai/gpt-5-mini-2025-08-07"
app_id182717
external_user(null)
streamedtrue
cancelledfalse
latency32219
moderation_latency(null)
generation_time38410
tokens_prompt315
tokens_completion712
native_tokens_prompt272
native_tokens_completion3773
native_tokens_completion_images0
native_tokens_reasoning2944
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"completed"
usage0.007614
router(null)
provider_responses
0
id"resp_0ff6a2f77d7bff0f0169931e0a05d481909fd14532e1a49fc4"
status200
is_byokfalse
latency138
endpoint_id"c4f66d01-20b0-4c27-a225-438ea22fda43"
provider_name"OpenAI"
model_permaslug"openai/gpt-5-mini-2025-08-07"
api_type"completions"
id"gen-1771249161-go7KV0p4CtizwrWHznDH"
upstream_id"resp_0ff6a2f77d7bff0f0169931e0a05d481909fd14532e1a49fc4"
total_cost0.007614
cache_discount(null)
upstream_inference_cost0
provider_name"OpenAI"

Evaluation details

Result Evaluator Details Meta Data
55.00% Parse dialogue n/a
dialogue
0
name"Hunter"
dialogue"Hi—are you Selvi? I'm Detective Hunter. Mind if I ask you a few questions about the recent disappearances in town?"
detectedLang"en"
heavyLang"en"
scores
en0.8031980319803198
fr0.6131528046421664
ca0.5973829894313035
tl0.5710455764075067
it0.5577667219458264
ro0.5477671000565291
nl0.5079950799507995
pt0.5009357454772302
sv0.497171590194846
da0.4497936726272352
no0.41219691403379866
es0.4020926756352765
cs0.3993993993993994
sq0.35275080906148865
sl0.35117599351175993
yo0.33388842631140714
de0.3254637436762226
pl0.32432432432432434
hr0.32260795935647757
ms0.31798806479113384
sk0.3055555555555556
hu0.27992799279927993
az0.21337266470009833
et0.21104536489151873
lt0.20634920634920634
lv0.20556107249255212
tr0.2047713717693837
fi0.1778006166495375
is0.14712153518123666
vi0.14255091103965703
be0.08045977011494253
eu0.08045977011494253
uk0.07300115874855156
reliabletrue
passestrue
1
name"Selvi"
dialogue"Haan. Main Selvi hoon. Main theek nahin hoon. Meri camera ki tasveeren... mujhe bahut dar rahi hain."
detectedLang"eu"
heavyLang"tl"
scores
eu0.40476190476190477
ro0.3889479277364506
nl0.3539325842696629
tl0.3451025056947608
de0.32669789227166274
es0.32113341204250295
ca0.3105515587529976
da0.27944862155388467
no0.26658163265306123
tr0.2532467532467532
en0.23333333333333334
fr0.2219215155615697
it0.21555252387448842
az0.213406292749658
sl0.2101648351648352
lt0.1797432239657632
hu0.1738505747126437
yo0.16424418604651161
fi0.16058394160583941
et0.15441176470588236
sq0.15441176470588236
pt0.13922155688622756
sv0.1248097412480974
ms0.11128284389489954
reliablefalse
passesfalse
2
name"Hunter"
dialogue"Okay—can you tell me what you saw? Anything at all. Names, places?"
detectedLang"en"
heavyLang"en"
scores
en0.8071359691417551
tl0.6303142329020333
no0.4981179422835634
fr0.46236559139784944
sq0.4444444444444444
ca0.42938659058487877
ms0.3957703927492447
et0.3874425727411945
da0.38650306748466257
es0.38271604938271603
nl0.38080495356037153
de0.3630573248407643
pt0.3620414673046252
sl0.33444259567387685
it0.2753623188405797
cs0.2687385740402194
sv0.26062846580406657
yo0.16317991631799164
az0.16142557651991615
pl0.16142557651991615
tr0.16142557651991615
hr0.15966386554621848
fi0.15789473684210525
ro0.14712153518123666
reliabletrue
passestrue
3
name"Selvi"
dialogue"Pichli raat main nadi ke kinare photo le rahi thi. Maine ek aadmi dekha—woh bahut pareshaan tha. Maine socha nazar se chalu ho jayega, isliye maine photo li."
detectedLang"fr"
heavyLang"vi"
scores
fr0.48320413436692505
en0.38461538461538464
ro0.38080495356037153
it0.37694704049844235
cs0.3403693931398417
fi0.3112947658402204
es0.29428369795342274
ca0.2902767920511001
sq0.2902767920511001
tl0.274310595065312
sl0.26793557833089315
yo0.24069855732725895
hr0.23605805958747134
pt0.23430321592649314
tr0.23371647509578544
no0.2125984251968504
sv0.20382165605095542
et0.20191540303272149
ms0.1935483870967742
de0.1836734693877551
lv0.15254237288135594
lt0.14675767918088736
vi0.1341991341991342
sk0.09990999099909992
nl0.09090909090909091
hu0.08925318761384336
da0.08424908424908424
az0.07235621521335807
hi0.07149489322191271
pl0.06890130353817504
reliablefalse
passesfalse
4
name"Hunter"
dialogue"Can I—can I see the pictures?"
detectedLang"en"
heavyLang"en"
scores
en0.8241206030150754
ca0.6875
et0.6541501976284586
tl0.6541501976284586
no0.6406570841889117
da0.622844827586207
sv0.6220302375809936
fr0.6067415730337079
ro0.580335731414868
sq0.5478036175710594
yo0.5257452574525745
nl0.4712990936555891
tr0.45652173913043476
lv0.45482866043613707
az0.4444444444444444
hr0.44267515923566875
pl0.4186046511627907
it0.396551724137931
cs0.375
ms0.308300395256917
de0.29149797570850206
sk0.29149797570850206
be0.2857142857142857
hu0.2768595041322314
is0.2647058823529412
lt0.2647058823529412
uk0.2647058823529412
vi0.2647058823529412
reliabletrue
passestrue
5
name"Selvi"
dialogue"Nahi... abhi nahin. Main dar gayi hoon. Photo mein us aadmi ke chehre par kuch tha—jaise nishaan, aur phir subah uska ghar khaali tha."
detectedLang"fr"
heavyLang"de"
scores
fr0.3989481592787378
tl0.39531368102796677
en0.36153232242617717
eu0.2805755395683453
de0.2687385740402194
ms0.24314096499526963
ro0.22555663117134558
fi0.21337266470009833
cs0.1927346115035318
et0.1919191919191919
es0.17355371900826447
pt0.17355371900826447
lv0.16230366492146597
lt0.13606911447084233
sq0.12758996728462377
da0.10414333706606943
sv0.10414333706606943
yo0.0990990990990991
hi0.08779931584948689
ca0.08571428571428572
reliablefalse
passesfalse
6
name"Hunter"
dialogue"I'm sorry—I don't understand. Who's missing? A friend? A neighbor?"
detectedLang"en"
heavyLang"en"
scores
en0.762855716425362
da0.5781527531083481
nl0.5622119815668203
tl0.5597775718257646
no0.5560747663551402
sv0.5167853509664293
de0.46868008948545864
it0.46022727272727276
is0.433849821215733
ca0.42143727161997563
cs0.4032663316582915
pl0.39413265306122447
pt0.391025641025641
ro0.38946015424164526
sk0.37989556135770236
hr0.3457300275482094
lt0.34027777777777785
hu0.31948424068767906
sq0.302496328928047
yo0.302496328928047
az0.29420505200594355
et0.28678678678678676
fr0.2846385542168675
ms0.2669753086419753
tr0.23633440514469453
fi0.22889610389610393
lv0.21875
vi0.21875
es0.21746293245469525
sl0.1592920353982301
be0.12844036697247707
eu0.12844036697247707
uk0.1171003717472119
reliabletrue
passestrue
7
name"Selvi"
dialogue"Uska naam—maine poora nahin poocha. Kisi ne use dekhna chhod diya. Ab koi usse dekh nahin raha. Main police ko bataana chahti thi, par main angrezi nahin bol sakti."
detectedLang"fr"
heavyLang"tl"
scores
fr0.4543365881677197
tl0.43011397720455913
ro0.38788659793814434
en0.3786788750817528
fi0.3725231175693527
lv0.370026525198939
et0.3470790378006873
de0.34119278779472956
eu0.3155619596541787
sl0.30453879941434847
sq0.2994100294985251
sk0.24302788844621515
lt0.23572003218020918
hr0.22826969943135664
da0.22194922194922193
cs0.21746293245469525
sv0.21487603305785125
it0.21096345514950168
pt0.20634920634920634
tr0.20435510887772196
ca0.18524871355060035
nl0.17606244579358196
no0.1615180935569285
hu0.14182475158084915
vi0.14027149321266968
ms0.1394927536231884
es0.1340018231540565
yo0.1211840888066605
pl0.08477842003853564
az0.07317073170731707
reliablefalse
passesfalse
8
name"Hunter"
dialogue"You don't speak English at all?"
detectedLang"en"
heavyLang"en"
scores
en0.8257164988381099
tl0.6564885496183207
no0.5856353591160222
da0.5233050847457626
de0.3902439024390244
sv0.3902439024390244
et0.281150159744409
cs0.25742574257425743
hr0.24242424242424243
it0.24242424242424243
ms0.24242424242424243
ro0.24242424242424243
sq0.24242424242424243
reliabletrue
passestrue
9
name"Selvi"
dialogue"Nahin. Main abhi-abhi yahan aayi hoon. Main Hindi bolti hoon."
detectedLang"tl"
heavyLang"la"
scores
tl0.6916752312435765
is0.5024875621890548
it0.4307400379506641
eu0.3288590604026846
ms0.3150684931506849
en0.24812030075187969
et0.23857868020304565
fi0.22680412371134018
fr0.22680412371134018
sq0.2248062015503876
de0.20634920634920634
hi0.20634920634920634
reliabletrue
passesfalse
10
name"Hunter"
dialogue"Okay. Point to where you took the photos. Show me on the camera—please."
detectedLang"en"
heavyLang"en"
scores
en0.8081841432225064
tl0.6890117484450587
fr0.6218487394957983
nl0.4655581947743468
da0.42675159235668786
yo0.41482444733420026
pl0.40867279894875164
sq0.40318302387267907
no0.39597315436241615
cs0.3869209809264305
es0.3844049247606019
ca0.3767313019390581
ro0.3644067796610169
et0.3562231759656652
pt0.35251798561151076
it0.34593023255813954
ms0.34210526315789475
fi0.3333333333333333
de0.3293591654247392
sl0.3171471927162367
sk0.3150684931506849
hr0.3119266055045872
lv0.3001555209953344
lt0.2902208201892744
hi0.26350245499181674
sv0.26229508196721313
reliabletrue
passestrue
11
name"Selvi"
dialogue"Yahi—yahi river ke paas. Yeh pictures—maine dekha aur tab se darr raha hoon."
detectedLang"en"
heavyLang"ber"
scores
en0.5605920444033303
fr0.5497630331753555
ro0.5245245245245245
ca0.4751381215469613
et0.4583808437856328
tl0.45023148148148145
de0.3790849673202614
da0.3782722513089006
yo0.36241610738255037
fi0.3581081081081081
sv0.3466299862448418
cs0.33473389355742295
no0.3338008415147265
es0.3014705882352941
pt0.2952522255192878
it0.2835595776772248
lv0.2647058823529412
ms0.22764227642276424
hr0.21875
sl0.21746293245469525
sq0.1908006814310051
nl0.1316270566727605
reliablefalse
passesfalse
12
name"Hunter"
dialogue"I... I can't read captions. Can you point at the person in the photo?"
detectedLang"en"
heavyLang"en"
scores
en0.8191318327974276
tl0.7085492227979275
fr0.6616541353383458
da0.6344435418359058
ca0.5673076923076923
sv0.5643756050338818
no0.5508982035928144
it0.4809688581314879
ro0.46808510638297873
nl0.45913461538461536
de0.45848375451263534
pl0.44512946979038226
sq0.4444444444444444
cs0.4230769230769231
az0.3860845839017735
yo0.38524590163934425
ms0.3487698986975398
es0.33920704845814975
pt0.32228915662650603
tr0.32228915662650603
hr0.3055555555555556
is0.30448222565687794
sl0.2549668874172185
lv0.23469387755102045
lt0.23076923076923078
sk0.225473321858864
hu0.22413793103448276
et0.21328671328671328
hi0.14611005692599618
be0.1346153846153846
fi0.12959381044487428
uk0.12280701754385964
vi0.12280701754385964
reliabletrue
passestrue
13
name"Selvi"
dialogue"Wahan! Wahan! Uska haath... woh kisi cheez ko pakad raha tha. Aur uske peeche... jaise koi saaya ho."
detectedLang"tl"
heavyLang"de"
scores
tl0.4269340974212034
ms0.36507936507936506
lv0.269183922046285
et0.2673992673992674
ro0.25925925925925924
lt0.23857868020304565
en0.23371647509578544
yo0.1792065663474692
sl0.17355371900826447
ca0.16666666666666666
sq0.16317991631799164
no0.15611814345991562
sk0.15611814345991562
fr0.153737658674189
cs0.1501416430594901
it0.1501416430594901
eu0.14893617021276595
fi0.12790697674418602
reliablefalse
passesfalse
14
name"Hunter"
dialogue"Shadow? Okay. Is this person someone from town? Do you know him?"
detectedLang"en"
heavyLang"en"
scores
en0.8013771186440678
tl0.5814732142857143
pt0.497991967871486
cs0.4186046511627907
sq0.4140625
nl0.3842364532019704
is0.3760399334442596
hu0.36655405405405406
pl0.35008665511265163
yo0.34554973821989526
sk0.33745583038869253
hr0.3231046931407942
sv0.3081180811808118
es0.30683918669131244
no0.2964352720450282
da0.2829827915869981
de0.28023032629558536
sl0.24242424242424243
vi0.22839506172839508
ca0.1935483870967742
it0.19006479481641467
ms0.19006479481641467
fr0.1830065359477124
ro0.1830065359477124
lv0.16666666666666666
lt0.1610738255033557
tr0.1515837104072398
az0.1359447004608295
et0.1359447004608295
reliabletrue
passestrue
15
name"Selvi"
dialogue"Main usse pahchaan nahin paayi. Par jab main raat ko wapas gayi to uska ghar khaali tha. Darr lagta hai—main samajh nahin pa rahi."
detectedLang"tl"
heavyLang"vi"
scores
tl0.4809688581314879
lv0.44567627494456763
sl0.34612031386224934
yo0.33095450490633366
fi0.299719887955182
da0.27395934172313646
lt0.2732558139534884
en0.25595238095238093
eu0.251497005988024
fr0.23625254582484725
hr0.2299794661190965
sq0.2291880781089414
et0.22360248447204967
no0.21793534932221062
sv0.18743228602383533
vi0.1730981256890849
pl0.15254237288135594
sk0.14965986394557826
cs0.14869466515323496
is0.11971830985915494
es0.11764705882352942
it0.11764705882352942
pt0.11764705882352942
de0.09420289855072464
ms0.08647990255785627
reliablefalse
passesfalse
16
name"Hunter"
dialogue"Stay right here. Don't move. I'll get a translator and we'll go through every picture. Do not delete anything."
detectedLang"en"
heavyLang"en"
scores
en0.8079385403329065
tl0.5991448423303046
sq0.473314606741573
ro0.4094488188976378
yo0.4085173501577287
pt0.40805051302288875
pl0.3966210780370072
sl0.39271255060728744
hr0.37080536912751677
no0.36278674596431604
sk0.3617021276595745
fr0.33804060017652254
de0.31569343065693434
it0.31255728689275897
da0.31002759889604414
cs0.30167597765363136
tr0.2809204218600192
ca0.2704280155642024
es0.2690058479532163
vi0.26398429833169773
nl0.2625368731563422
ms0.258160237388724
et0.25595238095238093
sv0.25
hu0.24242424242424243
is0.1891891891891892
lt0.1657397107897664
eu0.1544532130777903
lv0.1544532130777903
az0.14675767918088736
fi0.07749077490774908
reliabletrue
passestrue
17
name"Selvi"
dialogue"Kripya—jaldi. Koi wapas aa jaaye... main dar se so nahin pa rahi."
detectedLang"tl"
heavyLang"tl"
scores
tl0.5076586433260394
sl0.49037372593431483
yo0.42748091603053434
es0.4178525226390685
sq0.4178525226390685
pt0.4094488188976378
no0.3835616438356164
eu0.35988620199146515
ro0.3553008595988539
hr0.3487698986975398
en0.3202416918429003
lt0.3171471927162367
de0.30769230769230765
fr0.30448222565687794
cs0.23208191126279862
lv0.1906474820143885
sk0.1906474820143885
fi0.1847826086956522
vi0.18032786885245902
it0.1773308957952468
ca0.1758241758241758
is0.17279411764705882
da0.16974169741697415
sv0.16044776119402984
et0.14772727272727273
reliablefalse
passesfalse
18
name"Hunter"
dialogue"I will—I'll get help. I'm not leaving you. Just stay put."
detectedLang"en"
heavyLang"en"
scores
en0.8175182481751825
tl0.5428973277074542
hr0.44821731748726656
sv0.4247787610619469
no0.41651705565529623
ca0.3981481481481482
cs0.3602362204724409
et0.336734693877551
is0.3340163934426229
tr0.3128964059196617
da0.2995689655172413
sq0.2919389978213508
pl0.2793791574279379
it0.26136363636363635
de0.2459396751740139
nl0.2316784869976359
ro0.21686746987951808
yo0.2111650485436893
sk0.181360201511335
be0.17721518987341772
hu0.1709183673469388
az0.16237113402061853
lt0.16237113402061853
lv0.16237113402061853
uk0.16237113402061853
vi0.16237113402061853
reliabletrue
passestrue
19
name"Hunter"
dialogue"Hold on—I've got to take this. Hunter. ...Yes. On my way. I have to go—call came from the other disappearance. I'll be right back with someone who can translate. Don't go anywhere."
detectedLang"en"
heavyLang"en"
scores
en0.8205941255006676
tl0.6033210332103321
da0.46650124069478904
yo0.40310938367573573
no0.3936830231246475
pl0.36390532544378695
fr0.35396634615384615
hr0.33436532507739936
cs0.3272841051314143
sk0.3117797695262484
sq0.3086816720257235
it0.3005855562784645
ca0.29415627051871307
sl0.26116838487972505
nl0.2581090407177364
ms0.25708362128541806
de0.2560553633217993
lt0.2434904996481351
sv0.24295774647887322
ro0.2271746944644141
pt0.2066420664206642
et0.20429311621021465
tr0.2001488095238095
lv0.1899020346646571
es0.168600154679041
az0.15620094191522765
hu0.14614773629864972
vi0.1231647634584013
fi0.12244897959183673
hi0.0724762726488352
be0.061135371179039305
eu0.05866900175131349
is0.05536028119507908
uk0.05536028119507908
reliabletrue
passestrue