NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-01-01M1EKVKQYMTMW5G33ZFFCQMSP
Run Details
Model:
Grok 4.6 (Reasoning, High)
Test:
Bad Writing Habits
Scenario:
Mystery: examining a crime scene
Duration:
2m 32s
Cost:
$0.040112
Provider:
xAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Adverbs in dialogue tags
(details)
100.00%
AI-ism adverb frequency
(details)
100.00%
AI-ism character names
(details)
100.00%
AI-ism location names
(details)
80.60%
AI-ism word frequency
(details)
100.00%
Cliché density
(details)
100.00%
Emotion telling (show vs. tell)
(details)
100.00%
Filter word density
(details)
100.00%
Gibberish response detection
(details)
100.00%
Markdown formatting overuse
(details)
100.00%
Missing dialogue indicators (quotation marks)
(details)
82.27%
Name drop frequency
(details)
100.00%
Narrator intent-glossing
(details)
100.00%
"Not X but Y" pattern overuse
(details)
100.00%
Overuse of "that" (subordinate clause padding)
(details)
100.00%
Paragraph length variance
(details)
91.34%
Passive voice overuse
(details)
100.00%
Past progressive (was/were + -ing) overuse
(details)
100.00%
Em-dash & semicolon overuse
(details)
100.00%
Purple prose (modifier overload)
(details)
100.00%
Repeated phrase echo
(details)
100.00%
Sentence length variance
(details)
48.79%
Sentence opener variety
(details)
60.61%
Adverb-first sentence starts
(details)
100.00%
Pronoun-first sentence starts
(details)
32.73%
Subject-first sentence starts
(details)
0.00%
Subordinate conjunction sentence starts
(details)
100.00%
Technical jargon density
(details)
100.00%
Useless dialogue additions
(details)
100.00%
Dialogue tag variety (said vs. fancy)
(details)
89.8776%
Overall
Model input
System
You are an expert fiction writer. Always keep the following rules in mind: - Write in past tense and use General English spelling, grammar, and colloquialisms/slang. - Write in active voice - Always follow the "show, don't tell" principle. - Avoid adverbs and cliches and overused/commonly used phrases. Aim for fresh and original descriptions. - Convey events and story through dialogue. - Mix short, punchy sentences with long, descriptive ones. Drop fill words to add variety. - Skip "he/she said said" dialogue tags and convey people's actions or face expressions through their speech - Avoid mushy dialog and descriptions, have dialogue always continue the action, never stall or add unnecessary fluff. Vary the descriptions to not repeat yourself. - Put dialogue on its own paragraph to separate scene and action. - Reduce indicators of uncertainty like "trying" or "maybe" When writing text: - NEVER conclude the scene on your own, follow the beat instructions very closely. - NEVER end with foreshadowing. - NEVER write further than what I prompt you with. - AVOID imagining possible endings, NEVER deviate from the instructions. - STOP EARLY if the continuation contains what was required in the instructions. You do not need to fill out the full amount of words possible.
User/Human
Take into account the following glossary of characters/locations/items/lore... when writing your response: <codex> <character id="4416" name="Detective Harlow Quinn" aliases="Quinn, Detective Quinn, Harlow" pronouns="she/her" species="Human" storyRole="Antagonist"> <eyeColor>brown</eyeColor> <hair>closely cropped salt-and-pepper</hair> <height>5'9"</height> <age>41</age> <jawline>sharp jaw</jawline> <watch>worn leather watch on her left wrist</watch> <bearing>military precision</bearing> <description>A Metropolitan Police detective with 18 years of decorated service who suspects the clique is involved in criminal activity.</description> <backstory>She lost her partner, DS Morris, during a case three years ago under unexplained circumstances that had supernatural origins she doesn't yet understand.</backstory> </character> <character id="4414" name="Eva Kowalski" aliases="Eva, Evie" pronouns="she/her" species="Human" storyRole="Best Friend"> <eyeColor>green</eyeColor> <hair>curly red</hair> <height>5'4"</height> <age>26</age> <glasses>round glasses</glasses> <satchel>worn leather satchel full of books</satchel> <freckles>freckled complexion</freckles> <nervousHabit>tucking hair behind her left ear</nervousHabit> <description>Aurora's childhood friend and an occult researcher.</description> <backstory>Moved to London two years before Aurora and works as a research assistant at the British Museum's restricted archives. She holds a master's degree in Ancient History from Oxford.</backstory> </character> <location id="856" name="The Veil Market" aliases="Veil Market, the Market, underground market" realm="Earth"> <currentLocation>abandoned Tube station beneath Camden</currentLocation> <entryRequirement>bone token</entryRequirement> <schedule>moves locations every full moon</schedule> <description>A hidden supernatural black market that sells enchanted goods, banned alchemical substances, and information.</description> </location> <item id="2004" name="Veil Compass" aliases="the Compass, shadow compass" type="Tool" origin="The Veil Market"> <casingMaterial>small brass compass</casingMaterial> <faceMarkings>face etched with protective sigils</faceMarkings> <needle>points toward the nearest supernatural rift or portal</needle> <patina>casing has a patina of verdigris</patina> <creator>crafted by a Shade artisan</creator> <description>A compass attuned to supernatural energy.</description> </item> </codex> Continue the story using the following instructions: <instructions> Write a fully fleshed out scene of at least 1000 words in the mystery genre. A detective arrives at a crime scene in an unusual location. Something is wrong — the evidence doesn't add up. A colleague is present, offering a different interpretation. Through careful observation and deduction, the detective begins to see what others have missed. Use the codex entries above as your world reference for characters and setting. Only include the prose of the scene without any markdown or commentary. </instructions>
Model response
Messages
Meta
AI
The iron gate at the
Camden access tunnel hung open, its lock a twisted ruin on the concrete.
Detective Harlow Quinn stepped through
and into the dark.
She took the service stairs
two at a time, palm flat on the rail.
The worn leather of her
watch
caught on a flake of rust.
The air thickened as she
descended. Damp brick. Ozone.
A sweetness that belonged in
a confectioner's shop, not a gullet of tile and old iron.
A uniformed constable straightened at
the landing, torchlight jumping across his face. "Ma'am. Platform's this way. It's not a
standard
scene." She moved past him. The conversation ended without her. The station opened below.
An abandoned Tube platform, tracks
a dry trench of gravel and rat droppings.
Someone had raised a market
in the hollow.
Stalls of warped wood and
iron frames ran the length of the tiles, hung with canvas the colour of old bone.
Jars lined one table, contents
shifting behind the glass.
A rack of coats stood
beside a display of teeth strung on silver wire.
The electric lights the Met
had rigged threw a hard white across all of it.
The shadows behind the stalls
kept their own counsel. Yellow tape bisected the platform. Beyond it, a body. And Eva Kowalski.
The researcher from the British
Museum crouched near the dead man's feet, her worn leather satchel open on the tiles, round glasses catching the glare. Curly red hair fell across her freckled face.
She tucked it behind her
left ear without looking up. Quinn ducked the tape. "This is a crime scene, Miss Kowalski. Not an archive." Eva's green eyes lifted. "I was already here. I called your lot. Bone token got me through the gate before your constable even found the stairs." "You had a token." "I had a reason to be here. Restricted catalogues don't cover half of what moves through this place." Eva tipped her chin at the stalls. "Full moon was three nights back. Market should have shifted locations. It didn't. That alone is wrong." Quinn let her gaze travel the platform first, the way she had
been trained
, the way eighteen years had carved into muscle. Military lines. Corners. Exits. The body last. A man, forty or thereabouts, lay on his back between two stalls. Dark wool coat, collar open. A wound occupied the centre of his abdomen, the fabric stiff with blood. Around him, someone had poured a circle of white granules. Four candle stubs sat at the compass points of that circle, wax pooled but wicks unburnt. On his chest, a small brass compass, its casing mottled with verdigris, face
etched
in marks Quinn did not recognise. She crouched. The tiles under her knees held a film of dust broken by too many boots. SOCO had been and gone for the first sweep.
Their markers sat like little
yellow accusations beside a smear, a fibre, a print. "Name?" Eva drew a notebook from the satchel. "Stall ledger lists him as Corbin Vale. Shade artisan. He made tools. Attunement pieces. That compass is his work. Needle points toward the nearest rift." Quinn turned her wrist. The leather
watch
read 02:14. "Time of death?" "Your pathologist put it between eleven and midnight. I found him at half twelve. Market was empty. Stalls still up. Nobody shifts inventory and leaves a body in a salt circle for the police to trip over." Quinn studied the wound. The cut ran clean, a single thrust, no hesitation marks in the wool. Blood had soaked down into the coat and out across the tiles in a fan that stopped, neat as a drawn line, at the inner edge of the salt. "Blood doesn't respect circles, Miss Kowalski." Eva pushed her glasses up. "It does if the circle was laid first and the working held. Containment. Salt and iron filings. You can see the grey in it. Classic binding. Something came through, or something was sent back. Vale's compass would have led him to the weak point. He got here. He didn't leave." Quinn rose.
She walked the perimeter of
the salt without crossing it. The granules were uniform, white, grocery-fine.
She pinched a few between
finger and thumb and rubbed. No grit of iron. No
weight
. "Your filings aren't here." "They oxidise. They sink." "They don't vanish." Quinn let the salt fall.
She moved to the stall
on the left. Bottles. Dark glass, corked, labels in a script that hurt to look at too long. A gap in the middle shelf, a rectangle of cleaner wood where something had sat for months and then gone. Dust on every bottle except the ones at the front of that gap, which had
been handled
, shifted, set back a finger's width out of true. The stall on the right held trays of small bones and a cash box left open, notes still inside, coins untouched. "Thief who leaves the money." Quinn ran a gloved finger along the lip of the cash box. No blood. "And a killer who stages a ritual with supermarket salt and candles he never lights." Eva stood, satchel bumping her hip.
She was half a foot
shorter, and she had to tip her head back. The freckles across her nose stood out under the white lights. "You're reading it like a burglary that went sour. This market sells banned alchemical substances and information people kill for. Vale built compasses that find rifts. If a rift opened here, if something stepped through—" "Then I would have a body that walked here on its own feet." Quinn returned to the dead man.
She lifted the hem of
his coat with a pen. The backs of his shoes
were scored
, leather raw, gravel packed into the welt. The toes were clean. The platform dust around the salt circle showed a drag path, two parallel scuffs, interrupted where SOCO had planted a marker and walked through the rest. She looked at his face. Livor mortis stained the right cheek and the right side of the neck, a livid map that did not match a man found on his back.
He had lain on his
side for hours before someone rolled him. "Who moved him?" Eva tucked hair behind her ear again. "I didn't touch the body. I know better." "Someone did. He died on his side. Elsewhere. The abdomen wound was second. Look at the edges. Blood volume in the coat is wrong for a live bleed. You nick the aorta in a place like this and you get a pond, not a fan that stops at a line of salt. This blood was poured." Quinn lifted the compass with her pen through the bow. The needle swung,
trembled
, and settled.
It pointed at the brick
wall at the far end of the platform, where the tunnel mouth had
been blocked
decades ago, masonry paler than the surrounding brick, mortar still sandy at the joints. Eva's voice dropped. "That's where it opened. The needle doesn't lie. Vale followed it. Whatever came through killed him and went back. The circle was to hold it. It failed, or he failed to close it in time." "The needle points at fresh brickwork." Quinn walked toward the wall. Her heels clicked on tile, even,
measured
. She stopped a yard short. A scrape ran through the pale mortar, a crescent the width of a brass casing. Verdigris stained the groove, green on grey, a match for the compass in her gloved hand. Below the scrape, the dust held the square imprint of a crate, dragged, then lifted. No rift. A hole that someone had made and then bricked, and then opened again tonight. She turned the compass. The face markings caught the light. Protective sigils, Eva had called them. One of the
etched
lines had
been scored
over, a fresh scratch that broke the pattern and ran into the next. Not wear. A blade. Quinn looked back down the platform. The constable at the stairs. The stalls. The salt that was only salt. The candles with virgin wicks. The cash box full of notes. The gap on the shelf. The man who had not walked to his own death. Eva followed her gaze and then followed her, satchel swinging. "You think a person did this. A person with a crate and a knife and a working knowledge of how this market is supposed to look to someone like me." "I think your artisan is dead because he knew what was in that wall, and someone wanted it more than they wanted him breathing." Quinn crouched again, this time at the edge of the drag marks SOCO had truncated. A second set of prints sat in the dust beyond their markers, smaller than a man's, the tread a smooth oval, no heel.
They came from the blocked
tunnel and went back to it.
They did not approach the
body.
She set the compass on
the tile, outside the salt, and watched the needle.
It held on the pale
brick.
Then, as her shadow crossed
the face, it ticked one degree left, toward the gap on the bottle stall, toward the cleaner wood where a thing had sat and now did not. Eva went still. The glasses slipped down her nose and she left them. "The compass points to rifts." "Or to whatever it was attuned to last." Quinn's jaw worked.
She turned the dead man's
left wrist. No
watch
. A pale band of skin. On the platform beside his hip, half under the coat hem SOCO had missed, a bone token, still warm when she hovered her knuckles above it, warmer than a dead man's pocket had any right to be. She looked at Eva's satchel, at the mouth of it, at the books and the notebook and the edge of something pale tucked behind the spine of a volume on Mesopotamian seals. The token under the coat had a hairline crack. The one in the satchel, from this angle, did not. Quinn stood. The lights hummed. Somewhere in the dry tracks, water dripped with a patience older than the station. "You found him at half twelve." Eva's hand rose, stopped, and found her hair instead, tucking it back. "I did." "And you had a token to get in. The market that should have moved three nights ago. The artisan who makes compasses that find doors in walls. The crate pulled out of fresh brick. The wound that didn't kill him. The blood that was poured to make you see a circle and a working and a thing from the other side of a rift." Quinn's eyes, brown and flat in the white glare, held on the satchel. "What did he sell you, Miss Kowalski, before someone rolled him onto his back and wrote a ritual over a theft?"
created_at
"2026-09-01T13:52:06.919Z"
9/1/2026, 1:52:06 PM
model
"x-ai/grok-4.6-20260810"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
865
moderation_latency
(null)
generation_time
151662
tokens_prompt
1112
tokens_completion
3628
native_tokens_prompt
1204
native_tokens_completion
6316
native_tokens_completion_images
(null)
native_tokens_reasoning
4058
native_tokens_cached
128
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
service_tier
"default"
usage
0.040112
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788270726-3T9ieEcGfSiDIyCGh0r7"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788270726-N1NK8ljoZ85gp0Q8S0E3"
upstream_id
"6c239e0b-6516-9542-babb-c5d3d23dcba4"
provider_responses
0
endpoint_id
"0d0536e3-7eb1-4acb-8249-c8813365c2d8"
id
"6c239e0b-6516-9542-babb-c5d3d23dcba4"
is_byok
false
latency
864
model_permaslug
"x-ai/grok-4.6-20260810"
provider_name
"xAI"
status
200
total_cost
0.040112
cache_discount
0.000192
upstream_inference_cost
0
provider_name
"xAI"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Adverbs in dialogue tags
Target: ≤10% dialogue tags with adverbs
totalTags
7
adverbTagCount
1
adverbTags
0
"Quinn crouched again [again]"
dialogueSentences
32
tagDensity
0.219
leniency
0.438
rawRatio
0.143
effectiveRatio
0.063
100.00%
AI-ism adverb frequency
Target: <2% AI-ism adverbs (58 tracked)
wordCount
1804
totalAiIsmAdverbs
0
found
(empty)
highlights
(empty)
100.00%
AI-ism character names
Target: 0 AI-default names (17 tracked, −20% each)
codexExemptions
(empty)
found
(empty)
100.00%
AI-ism location names
Target: 0 AI-default location names (33 tracked, −20% each)
codexExemptions
(empty)
found
(empty)
80.60%
AI-ism word frequency
Target: <2% AI-ism words (290 tracked)
wordCount
1804
totalAiIsms
7
found
0
word
"standard"
count
1
1
word
"electric"
count
1
2
word
"etched"
count
2
3
word
"weight"
count
1
4
word
"trembled"
count
1
5
word
"measured"
count
1
highlights
0
"standard"
1
"electric"
2
"etched"
3
"weight"
4
"trembled"
5
"measured"
100.00%
Cliché density
Target: ≤1 cliche(s) per 800-word window
totalCliches
0
maxInWindow
0
found
(empty)
highlights
(empty)
100.00%
Emotion telling (show vs. tell)
Target: ≤3% sentences with emotion telling
emotionTells
0
narrationSentences
126
matches
(empty)
100.00%
Filter word density
Target: ≤3% sentences with filter/hedge words
filterCount
2
hedgeCount
0
narrationSentences
126
filterMatches
0
"watch"
hedgeMatches
(empty)
100.00%
Gibberish response detection
Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words)
analyzedSentences
151
gibberishSentences
0
adjustedGibberishSentences
0
longSentenceCount
0
runOnParagraphCount
0
giantParagraphCount
0
wordSaladCount
0
repetitionLoopCount
0
controlTokenCount
0
repeatedSegmentCount
0
maxSentenceWordsSeen
77
ratio
0
matches
(empty)
100.00%
Markdown formatting overuse
Target: ≤5% words in markdown formatting
markdownSpans
0
markdownWords
0
totalWords
1804
ratio
0
matches
(empty)
100.00%
Missing dialogue indicators (quotation marks)
Target: ≤10% speech attributions without quotation marks
totalAttributions
2
unquotedAttributions
0
matches
(empty)
82.27%
Name drop frequency
Target: ≤1.0 per-name mentions per 100 words
totalMentions
38
wordCount
1255
uniqueNames
10
maxNameDensity
1.35
worstName
"Quinn"
maxWindowNameDensity
2.5
worstWindowName
"Quinn"
discoveredNames
Camden
1
Harlow
1
Quinn
17
Tube
1
Met
1
Eva
13
Kowalski
1
British
1
Museum
1
Mesopotamian
1
persons
0
"Harlow"
1
"Quinn"
2
"Met"
3
"Eva"
4
"Kowalski"
5
"Museum"
6
"Mesopotamian"
places
0
"Camden"
1
"British"
globalScore
0.823
windowScore
0.833
100.00%
Narrator intent-glossing
Target: ≤2% narration sentences with intent-glossing patterns
analyzedSentences
81
glossingSentenceCount
0
matches
(empty)
100.00%
"Not X but Y" pattern overuse
Target: ≤1 "not X but Y" per 1000 words
totalMatches
0
per1kWords
0
wordCount
1804
matches
(empty)
100.00%
Overuse of "that" (subordinate clause padding)
Target: ≤2% sentences with "that" clauses
thatCount
0
totalSentences
151
matches
(empty)
100.00%
Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs
56
mean
32.21
std
24.24
cv
0.752
sampleLengths
0
27
1
51
2
12
3
9
4
9
5
93
6
9
7
3
8
43
9
4
10
10
11
26
12
4
13
42
14
29
15
76
16
40
17
1
18
32
19
12
20
37
21
46
22
6
23
55
24
34
25
4
26
4
27
16
28
61
29
21
30
34
31
66
32
13
33
60
34
43
35
3
36
15
37
56
38
50
39
38
40
23
41
62
42
41
43
45
44
40
45
79
46
52
47
13
48
5
49
62
91.34%
Passive voice overuse
Target: ≤2% passive sentences
passiveCount
5
totalSentences
126
matches
0
"been trained"
1
"been handled"
2
"were scored"
3
"been blocked"
4
"been scored"
100.00%
Past progressive (was/were + -ing) overuse
Target: ≤2% past progressive verbs
pastProgressiveCount
0
totalVerbs
178
matches
(empty)
100.00%
Em-dash & semicolon overuse
Target: ≤2% sentences with em-dashes/semicolons
emDashCount
0
semicolonCount
0
flaggedSentences
0
totalSentences
151
ratio
0
matches
(empty)
100.00%
Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount
1256
adjectiveStacks
0
stackExamples
(empty)
adverbCount
20
adverbRatio
0.01592356687898089
lyAdverbCount
2
lyAdverbRatio
0.0015923566878980893
100.00%
Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences
151
echoCount
0
echoWords
(empty)
100.00%
Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences
151
mean
11.95
std
11.31
cv
0.946
sampleLengths
0
18
1
9
2
14
3
12
4
6
5
2
6
1
7
16
8
12
9
9
10
4
11
5
12
4
13
13
14
8
15
21
16
9
17
14
18
15
19
9
20
5
21
4
22
3
23
25
24
8
25
10
26
4
27
10
28
4
29
22
30
4
31
25
32
17
33
22
34
2
35
1
36
1
37
3
38
12
39
5
40
13
41
10
42
16
43
20
44
2
45
15
46
9
47
14
48
1
49
7
48.79%
Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats
18
diversityRatio
0.36423841059602646
totalSentences
151
uniqueOpeners
55
60.61%
Adverb-first sentence starts
Target: ≥3% sentences starting with an adverb
adverbCount
2
totalSentences
110
matches
0
"Then, as her shadow crossed"
1
"Somewhere in the dry tracks,"
ratio
0.018
100.00%
Pronoun-first sentence starts
Target: ≤30% sentences starting with a pronoun
pronounCount
21
totalSentences
110
matches
0
"She took the service stairs"
1
"She moved past him."
2
"She tucked it behind her"
3
"Their markers sat like little"
4
"She walked the perimeter of"
5
"She pinched a few between"
6
"She moved to the stall"
7
"She was half a foot"
8
"She lifted the hem of"
9
"She looked at his face."
10
"He had lain on his"
11
"It pointed at the brick"
12
"Her heels clicked on tile,"
13
"She stopped a yard short."
14
"She turned the compass."
15
"They came from the blocked"
16
"They did not approach the"
17
"She set the compass on"
18
"It held on the pale"
19
"She turned the dead man's"
ratio
0.191
32.73%
Subject-first sentence starts
Target: ≤72% sentences starting with a subject
subjectCount
94
totalSentences
110
matches
0
"The iron gate at the"
1
"Detective Harlow Quinn stepped through"
2
"She took the service stairs"
3
"The worn leather of her"
4
"The air thickened as she"
5
"A sweetness that belonged in"
6
"A uniformed constable straightened at"
7
"She moved past him."
8
"The conversation ended without her."
9
"The station opened below."
10
"An abandoned Tube platform, tracks"
11
"Someone had raised a market"
12
"Stalls of warped wood and"
13
"Jars lined one table, contents"
14
"A rack of coats stood"
15
"The electric lights the Met"
16
"The shadows behind the stalls"
17
"The researcher from the British"
18
"She tucked it behind her"
19
"Quinn ducked the tape."
ratio
0.855
0.00%
Subordinate conjunction sentence starts
Target: ≥2% sentences starting with a subordinating conjunction
subConjCount
0
totalSentences
110
matches
(empty)
ratio
0
100.00%
Technical jargon density
Target: ≤6% sentences with technical-jargon patterns
analyzedSentences
53
technicalSentenceCount
0
matches
(empty)
100.00%
Useless dialogue additions
Target: ≤5% dialogue tags with trailing filler fragments
totalTags
7
uselessAdditionCount
0
matches
(empty)
100.00%
Dialogue tag variety (said vs. fancy)
Target: ≤10% fancy dialogue tags
totalTags
0
89.8776%