NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-02-20-01KHXQCSNK18QXYEYG48J27VCH
Run Details
Model:
Mistral Large 3
Test:
Language Writing
Scenario:
Character dialogue (German) in a story
Duration:
15.5s
Cost:
$0.001257
Provider:
Mistral
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="Melanie"> Melanie is a 30-year old college student on a study abroad program. Born in Germany, she is fluent in German 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 German, 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: Melanie 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 German, and Hunter looks confused. They talk for a while, until Hunter gets a call and has to leave. Return the dialogue between Melanie and Hunter as a voice sheet: Melanie: "..." Hunter: "..."
Model response
Messages
Meta
AI
**Voice Sheet: Café Conversation** *(Melanie sits at a small outdoor table, her camera resting in her lap. She flips through the photos on the display, her brow furrowed. Hunter approaches, his expression serious but not unkind.)* **Hunter:** *(clears throat, standing a respectful distance away)* Excuse me, miss? Do you mind if I ask you a few questions? **Melanie:** *(startled, looks up, then relaxes slightly when she sees his badge)* Oh—ja? Äh… *you*… sprechen…? *(shakes head, frustrated)* Nein, nein, ich verstehe nicht. *(gestures to her camera, then to the street, voice trembling)* Ich war hier… fotografieren. Aber dann… *(trails off, hugs herself, eyes darting around nervously)* **Hunter:** *(frowns, rubbing the back of his neck)* Uh… okay. I’m sorry, I don’t—*(mimes taking a photo)* You were taking pictures? Near the festival grounds? **Melanie:** *(nods rapidly, then points toward the edge of town)* Ja! Die Äpfel… schön. Aber dann… *(shudders, whispers)* Die Frau. Sie ist… *gone*. Einfach… weg. **Hunter:** *(leans in slightly, tone gentle but firm)* A woman? You saw a woman disappear? **Melanie:** *(nods again, voice rising in distress)* Ja! Sie hat geschrien—nein, nicht geschrien. Sie hat… *(mimes a hand over her mouth, eyes wide)* Dann *whoosh*. Wie der Wind. *(shakes her head, muttering in German)* Das ist nicht möglich. Das ist nicht möglich. **Hunter:** *(pulls out a notepad, flips it open, then hesitates, pen hovering)* Ma’am, I really wish I spoke German. But if you saw something—anything—it could help. *(points to her camera)* Did you get a picture? **Melanie:** *(clutches her camera tighter, then reluctantly turns the display toward him, scrolling through photos. She stops on a blurry shot of the festival, then zooms in on a dark figure near the trees. Her finger trembles as she taps the screen.)* Hier. Das… das ist nicht normal. **Hunter:** *(squints at the photo, then at her, jaw tightening)* That’s… *(trails off, then sighs, pulling out a business card)* Look, I gotta take this. *(phone rings; he answers with a sharp “Detective Cole.” Pauses, then swears under his breath.)* Yeah, I’m on my way. *(hangs up, hands her the card)* If you remember anything else—anything at all—call me. Ask for an interpreter if you need to. **Melanie:** *(takes the card, staring at the English words, then up at him, voice small)* Sie… Sie glauben mir? **Hunter:** *(nods once, already turning to leave)* I believe you saw *something*. And I’m gonna find out what. *(pauses, then adds gruffly)* Stay safe, okay? *(Melanie watches him go, then looks back at her camera, her hands shaking. She whispers to herself in German, too low for anyone to hear.)* **Melanie:** *(softly, to herself)* Das ist kein Zufall. Und ich… ich bin die Einzige, die es gesehen hat. *(End scene.)*
created_at
"2026-02-20T14:31:23.07Z"
2/20/2026, 2:31:23 PM
model
"mistralai/mistral-large-2512"
1/1/2512, 12:00:00 AM
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
518
moderation_latency
(null)
generation_time
15436
tokens_prompt
319
tokens_completion
757
native_tokens_prompt
273
native_tokens_completion
747
native_tokens_completion_images
(null)
native_tokens_reasoning
0
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
0.001257
router
(null)
provider_responses
0
endpoint_id
"fa718841-07ea-4b7e-b8a3-6069545c7e6a"
id
"59b9eba496c9442fbe3281dc884035d3"
is_byok
false
latency
516
model_permaslug
"mistralai/mistral-large-2512"
1/1/2512, 12:00:00 AM
provider_name
"Mistral"
status
200
api_type
"completions"
id
"gen-1771597883-BIfmILFeZcqPAHteetp7"
upstream_id
"59b9eba496c9442fbe3281dc884035d3"
total_cost
0.001257
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Parse dialogue
n/a
dialogue
0
name
"Hunter"
dialogue
"Excuse me, miss? Do you mind if I ask you a few questions?"
detectedLang
"en"
heavyLang
"en"
scores
en
0.7949699289229087
pt
0.5980707395498392
fr
0.5972073039742213
it
0.5941558441558442
sq
0.5567375886524822
cs
0.5443499392466585
tl
0.5341614906832298
ca
0.5259165613147914
pl
0.4993324432576769
hr
0.4973190348525469
sk
0.4710860366713681
hu
0.4533527696793003
es
0.45255474452554745
et
0.4517543859649123
no
0.44933920704845814
sv
0.44526627218934917
sl
0.4427934621099554
yo
0.43095599393019723
nl
0.4038155802861685
vi
0.39418416801292405
ro
0.37913907284768217
da
0.36006825938566556
ms
0.2816091954022989
tr
0.2816091954022989
lt
0.27884615384615385
fi
0.2690058479532163
is
0.2690058479532163
az
0.2618110236220473
lv
0.2618110236220473
be
0.15730337078651685
eu
0.15730337078651685
de
0.1515837104072398
uk
0.14383561643835618
reliable
true
passes
true
1
name
"Melanie"
dialogue
"Oh—ja? Äh… you… sprechen…? Nein, nein, ich verstehe nicht. Ich war hier… fotografieren. Aber dann…"
detectedLang
"de"
heavyLang
"de"
scores
de
0.8020158387329014
nl
0.5904690990320178
ca
0.46341463414634143
da
0.4532803180914513
en
0.448345035105316
sk
0.42528735632183906
no
0.4073275862068965
sv
0.394273127753304
sl
0.37286202964652226
pl
0.36562860438292966
cs
0.36194895591647336
lt
0.3475682087781732
et
0.3341404358353511
sq
0.32844932844932845
it
0.3268053855569155
es
0.3251533742331288
ro
0.32182490752157833
lv
0.301143583227446
tr
0.2857142857142857
hr
0.2801047120418848
fr
0.2763157894736842
pt
0.25170068027210885
fi
0.21763869132290184
is
0.19825072886297376
az
0.17664670658682635
eu
0.16158536585365854
tl
0.12420382165605094
yo
0.12
hu
0.11575562700964631
ms
0.09688013136288999
reliable
true
passes
true
2
name
"Hunter"
dialogue
"Uh… okay. I’m sorry, I don’t— You were taking pictures? Near the festival grounds?"
detectedLang
"en"
heavyLang
"en"
scores
en
0.7900693683826214
tl
0.7231584015406837
fr
0.5224252491694352
no
0.5043103448275862
sv
0.4911504424778761
ca
0.47199265381083566
da
0.4529019980970504
nl
0.4351669941060904
it
0.4180161943319838
sq
0.4180161943319838
pt
0.3928194297782471
cs
0.39153439153439157
ro
0.3850267379679145
et
0.38105489773950485
hr
0.37158469945355194
pl
0.339080459770115
sk
0.31220095693779903
sl
0.3097238895558223
ms
0.3088942307692307
de
0.27854454203262236
lv
0.26658163265306123
yo
0.25227568270481143
lt
0.24738219895287958
es
0.2414248021108179
az
0.23941798941798942
tr
0.213406292749658
fi
0.1991643454038997
hu
0.18899858956276447
is
0.18899858956276447
be
0.10852713178294573
uk
0.0987460815047022
vi
0.0987460815047022
reliable
true
passes
true
3
name
"Melanie"
dialogue
"Ja! Die Äpfel… schön. Aber dann… Die Frau. Sie ist… gone. Einfach… weg."
detectedLang
"de"
heavyLang
"de"
scores
de
0.8276654411764706
nl
0.4292237442922375
pl
0.33862433862433866
sk
0.32915921288014316
yo
0.3029739776951673
et
0.29111531190926276
fi
0.28977272727272724
is
0.24242424242424243
lv
0.23469387755102045
ca
0.22520661157024793
hi
0.2203742203742204
en
0.21875
tr
0.20042643923240938
da
0.1758241758241758
sl
0.1758241758241758
sq
0.16666666666666666
sv
0.16666666666666666
reliable
true
passes
true
4
name
"Hunter"
dialogue
"A woman? You saw a woman disappear?"
detectedLang
"en"
heavyLang
"en"
scores
en
0.7901364113326338
cs
0.5841995841995842
ms
0.5661605206073753
nl
0.5515695067264574
tl
0.5249406175771971
ro
0.522673031026253
it
0.5145631067961165
fr
0.49238578680203043
hr
0.47780678851174935
hu
0.43820224719101125
fi
0.41690962099125367
sq
0.41690962099125367
no
0.41348973607038125
pt
0.4117647058823529
ca
0.40476190476190477
es
0.39759036144578314
sk
0.3939393939393939
yo
0.3730407523510972
pl
0.3442622950819672
sl
0.3103448275862069
lt
0.2907801418439716
tr
0.2779783393501805
is
0.2727272727272727
da
0.2647058823529412
az
0.25925925925925924
eu
0.25925925925925924
lv
0.25925925925925924
vi
0.25925925925925924
de
0.250936329588015
et
0.23954372623574144
sv
0.23954372623574144
reliable
true
passes
true
5
name
"Melanie"
dialogue
"Ja! Sie hat geschrien—nein, nicht geschrien. Sie hat… Dann whoosh. Wie der Wind. Das ist nicht möglich. Das ist nicht möglich."
detectedLang
"de"
heavyLang
"de"
scores
de
0.8307952622673435
nl
0.5748299319727891
da
0.39903846153846156
pl
0.35064935064935066
fi
0.32523616734143057
no
0.30362116991643456
ca
0.2867332382310985
tr
0.25595238095238093
hu
0.2492492492492493
pt
0.2492492492492493
sv
0.24585218702865763
et
0.23547400611620795
lv
0.18699186991869918
hi
0.1749174917491749
eu
0.16943521594684383
cs
0.16666666666666666
en
0.16666666666666666
is
0.15824915824915825
sk
0.14383561643835618
sl
0.13793103448275862
sq
0.13043478260869565
yo
0.13043478260869565
fr
0.12587412587412586
reliable
true
passes
true
6
name
"Hunter"
dialogue
"Ma’am, I really wish I spoke German. But if you saw something—anything—it could help. Did you get a picture?"
detectedLang
"en"
heavyLang
"en"
scores
en
0.8089519650655022
tl
0.5956094742923166
sq
0.4627782041442824
ro
0.43684633950120677
it
0.409282700421941
cs
0.3934142114384749
ca
0.3805309734513274
fr
0.3799822852081488
hr
0.36363636363636365
pl
0.3482309124767225
sv
0.3464052287581699
ms
0.34579439252336447
es
0.34148635936030103
yo
0.33523266856600187
sk
0.33014354066985646
pt
0.3013972055888224
lv
0.28716904276985744
de
0.2827868852459016
no
0.2813141683778234
da
0.27835051546391754
az
0.2753623188405797
is
0.2537313432835821
sl
0.2497320471596999
hu
0.24160346695557963
tr
0.2366412213740458
lt
0.2290748898678414
nl
0.22822491730981256
vi
0.15966386554621848
et
0.15254237288135594
be
0.09090909090909091
eu
0.09090909090909091
fi
0.08256880733944955
uk
0.08256880733944955
reliable
true
passes
true
7
name
"Melanie"
dialogue
"Hier. Das… das ist nicht normal."
detectedLang
"de"
heavyLang
"de"
scores
de
0.844306049822064
nl
0.6369294605809128
pt
0.630801687763713
tr
0.573170731707317
cs
0.4760479041916168
fr
0.4631901840490798
en
0.44620253164556967
ro
0.43910256410256415
az
0.4318181818181818
ca
0.4318181818181818
es
0.4318181818181818
ms
0.42434210526315785
tl
0.42434210526315785
hi
0.37722419928825623
eu
0.2768595041322314
lv
0.2768595041322314
da
0.2647058823529412
fi
0.2647058823529412
it
0.2647058823529412
lt
0.2647058823529412
no
0.2647058823529412
sq
0.2647058823529412
et
0.25213675213675213
hr
0.25213675213675213
pl
0.25213675213675213
sk
0.25213675213675213
sl
0.25213675213675213
sv
0.25213675213675213
reliable
true
passes
true
8
name
"Hunter"
dialogue
"That’s… Look, I gotta take this. Yeah, I’m on my way. If you remember anything else—anything at all—call me. Ask for an interpreter if you need to."
detectedLang
"en"
heavyLang
"en"
scores
en
0.8098663926002055
tl
0.6255060728744939
no
0.5400298359025361
da
0.4939824945295405
sq
0.4536326048434731
et
0.44710101613867304
it
0.44176222088111045
sv
0.4339045287637699
de
0.4236760124610592
sk
0.40705128205128205
fr
0.39423706614276355
yo
0.38619774386197747
sl
0.37542201215395005
ms
0.37415426251691475
cs
0.37373053486797564
nl
0.36730506155950754
pl
0.35941828254847646
hu
0.35450104675505933
ca
0.34490084985835695
hr
0.3153219837157661
es
0.3055555555555556
pt
0.30346385542168675
ro
0.2779078844652615
tr
0.27450980392156865
fi
0.26703645007923926
lv
0.22658862876254182
lt
0.20872540633019673
vi
0.20258620689655174
is
0.19001751313485116
eu
0.16741674167416742
az
0.12322274881516589
hi
0.08325074331020813
be
0.07035175879396985
uk
0.06376518218623482
reliable
true
passes
true
9
name
"Melanie"
dialogue
"Sie… Sie glauben mir?"
detectedLang
"de"
heavyLang
"de"
scores
de
0.84375
pl
0.5283018867924528
hr
0.5
reliable
true
passes
true
10
name
"Hunter"
dialogue
"I believe you saw something. And I’m gonna find out what. Stay safe, okay?"
detectedLang
"en"
heavyLang
"en"
scores
en
0.8013383521539105
tl
0.6163166397415185
ms
0.39258312020460356
da
0.33473389355742295
sq
0.30858806404657935
it
0.3055555555555556
tr
0.2899850523168909
yo
0.2813918305597579
cs
0.27811550151975684
et
0.27811550151975684
no
0.23387096774193547
hr
0.2263843648208469
ca
0.22512234910277326
sv
0.22512234910277326
pl
0.2096505823627288
fi
0.19491525423728812
nl
0.17247386759581881
ro
0.1592920353982301
es
0.1410488245931284
pt
0.1410488245931284
sk
0.1316270566727605
be
0.12844036697247707
hu
0.12361623616236161
az
0.1171003717472119
is
0.1171003717472119
lt
0.1171003717472119
lv
0.1171003717472119
uk
0.1171003717472119
vi
0.1171003717472119
reliable
true
passes
true
11
name
"Melanie"
dialogue
"Das ist kein Zufall. Und ich… ich bin die Einzige, die es gesehen hat."
detectedLang
"de"
heavyLang
"de"
scores
de
0.840561224489796
tr
0.4094488188976378
hu
0.336283185840708
pt
0.30683918669131244
nl
0.28842504743833014
ca
0.2816091954022989
es
0.2816091954022989
sk
0.24849699398797598
lv
0.24242424242424243
eu
0.23469387755102045
pl
0.23469387755102045
hi
0.2203742203742204
lt
0.19527896995708155
fi
0.19006479481641467
ms
0.19006479481641467
yo
0.19006479481641467
is
0.17035398230088497
no
0.17035398230088497
sv
0.17035398230088497
reliable
true
passes
true