NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M37Q2MVSBPGBAEBTD6M293RP
Run Details
Model:
Aion 3.5 (Reasoning, High)
Test:
Bad Writing Habits
Scenario:
Mystery: examining a crime scene
Duration:
48.5s
Cost:
$0.015716
Provider:
AionLabs
Evaluation summary
Score
Evaluator
Details Link
100.00%
Adverbs in dialogue tags
(details)
92.10%
AI-ism adverb frequency
(details)
100.00%
AI-ism character names
(details)
100.00%
AI-ism location names
(details)
92.10%
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)
50.00%
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)
100.00%
Passive voice overuse
(details)
91.60%
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)
90.14%
Sentence opener variety
(details)
53.76%
Adverb-first sentence starts
(details)
100.00%
Pronoun-first sentence starts
(details)
56.77%
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)
90.8825%
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 ladder down into the
old station ran forty feet, and every rung rang like a struck bell in the dark.
Detective Harlow Quinn counted them
out of habit. Forty rungs, then her boots found the platform, and the smell hit her: rust, wet concrete, and underneath it something colder, like the air over a frozen pond. "Down here, ma'am."
A constable's torch beam wobbled
at the far end of the platform. Quinn walked.
Her torch swept the curved
tunnel walls, tiles the colour of old teeth, a roundel sign so grimed with decades that only the ghost of the station name survived: CAMDEN.
The track bed had been
drained years ago, but water still wept down the walls in long black streaks.
The body lay in the
mouth of the northbound tunnel. Male, fifties, grey suit that had cost more than her monthly salary. One arm flung across the sleepers, legs straight, eyes open toward the ceiling. No blood. No wounds she could see from three feet away. DS Iyer crouched beside him, gloved hands hovering over the chest. Young, sharp, still at the stage where he narrated everything he did. "Hypothermia, most likely," Iyer said. "Doc puts time of death between one and three this morning. Body temp's consistent with cold exposure, and it's bloody freezing down here. He's a rough sleeper, dressed nice for one, but rough sleepers end up in odd kit all the time. Wandered in through the maintenance access on the surface, got lost in the dark, sat down, and that was that." Quinn crouched opposite him.
The dead man's shoes were
Oxford leather, soles barely scuffed.
She pressed the back of
her glove to his cheek. Cold. Too cold. "It's June," she said. "Temperature drops like a stone in these deep tunnels. No airflow, the concrete pulls the heat out of—" "Sixteen degrees down here. I checked the surface access on my way in. It's a mild night."
She lifted the man's hand
.
The fingers were stiff, and
something white showed between them.
She worked it free with
tweezers and held it to the torchlight. A token. Carved from bone, the size of a pound coin, a hole bored through the middle.
Symbols ringed its edge, worn
smooth by handling. Iyer leaned in. "Domino? Some kind of game piece?" "Dominoes aren't carved from bone with sigils on them."
She bagged it
. "And rough sleepers don't freeze to death clutching them." "Maybe he collected oddities. You get magpie types."
Iyer stood, brushing grit from
his knees. "Honestly, ma'am, the simplest read is exposure. Doc agrees. There's no sign of assault, no defensive wounds, wallet's gone but that's scavengers, or he dropped it. Tragic, not suspicious." "Then explain the dust." Iyer blinked. "Sorry?"
Quinn swung her torch low
across the track bed.
The gravel between the sleepers
carried a fine grey silt, undisturbed, a smooth grey carpet running the length of the tunnel.
The dead man's shoe prints
started three metres from the body. She walked back along them, Iyer trailing behind, and the prints simply stopped. Two clean impressions, heel and toe, and then nothing. No drag marks. No scuffing. Just silt. "He walked to that spot," Quinn said, "and then he was sitting where he is now. How does a man walk three metres without touching the ground?" "Kids come down here. Urban explorers. They could have—" "Trampled it everywhere. They didn't."
She crouched at the last
footprint. "These prints don't approach the body. They leave it. Direction of travel's readable from the tread — he was walking toward the platform." Iyer's jaw worked. "That's not possible." "No. It isn't." A voice
echoed
down the tunnel from the platform, high and breathless. "Hello? Detective Quinn? It's Eva — Eva Kowalski? From the museum? Your office rang the archives at some ungodly hour and said you needed—" "On the track bed, Miss Kowalski. Mind the gap." Boots crunched on gravel and a small figure picked her way toward them, curly red hair escaping under a hard hat, a leather satchel banging against her hip.
She had a torch strapped
to the satchel's strap like a caving lamp, round glasses fogging at the edges. Eva stopped at the body and went
very
still. Then she looked past it, up the tunnel, into the dark. "You found him here?"
she said quietly
. "Who's this?" Iyer asked. "Consultant," Quinn said. "Miss Kowalski works restricted archives at the British Museum. My office sent photographs of the token down to the antiquities desk this evening. She recognised the carving." "Recognised is generous." Eva pulled a folded printout from her satchel and opened it against her knee: a photograph of a clay tablet, cuneiform and something older, and along its lower edge the same ring of symbols as the bone token. "This is a warding script. Pre-Roman, possibly older, and the reason it's in a restricted archive is that every documented example was found at sites with unexplained deaths. The script marks thresholds. Boundaries. Places where the world is thin."
She tucked a curl behind
her left ear. "It's not a game piece, and whoever carved it wasn't playing. It's a key." "A key to what?" Iyer's laugh came out thin. "It's a disused Tube station." Eva didn't answer.
She was staring at the
wall. Quinn followed her gaze. The tunnel wall here ran with condensation, except for one patch, a metre across, dry as chalk.
She crossed to it and
laid her palm flat against the tiles. Cold bit through her glove, sharp enough to make her hiss. Frost feathered out from her fingers. "It's June," Iyer said again, weaker this time. Quinn searched the dead man's pockets, working methodically while Eva's lamp lit the body. Inside pocket: a return ticket to Camden, dated today, unused. Trouser pocket: a small brass compass on a chain, its casing green with verdigris, its face
etched
with the same symbols as the token. The needle
was spinning
. Not swinging. Not hunting. Spinning, smooth and constant, like a fan blade. "That's broken," Iyer said. Quinn held the compass level and turned a slow circle. The needle ignored her, ignored north, ignored every direction — except one. Each time she rotated the casing, the needle swung back to the same bearing. She sighted along it. The dry, frosted patch of wall. "He came down here with a ticket he never used and a key he never spent," Quinn said. "And these prints don't approach the body because he didn't walk out. Something put him out."
She looked at the frost-rimed
tiles, then at the token in its evidence bag, and the etchings matched, symbol for symbol, the sigils on the compass face. "He was turned away at a door we can't see. And tonight's a full moon." Eva's head came up
sharply
. "How did you know that matters?" "Three years ago my partner died in a cellar in Whitechapel with no wounds, no toxins, and a room temperature of thirty degrees." Quinn kept her eyes on the wall, on the frost crawling in slow ferns across the tiles. "The report said cardiac arrest. The report was wrong. He had a bone token in his hand too, and nobody would tell me what it was."
She turned the compass over
in her gloved palm, watching the needle strain toward the wall like a dog at the end of a lead. "I stopped believing in coincidence around the same time I stopped believing in that report. This man wasn't lost, Iyer. He was a customer. And whatever market he was shopping at doesn't give refunds."
created_at
"2026-09-23T18:05:05.804Z"
9/23/2026, 6:05:05 PM
model
"aion-labs/aion-3.5-20260923"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
711
moderation_latency
(null)
generation_time
48428
tokens_prompt
1112
tokens_completion
2773
native_tokens_prompt
1019
native_tokens_completion
2492
native_tokens_completion_images
(null)
native_tokens_reasoning
915
native_tokens_cached
1019
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
"stop"
service_tier
(null)
usage
0.01571625
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790186705-TdG8ngdDdjb3OkQUGqMS"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790186705-uA5GnBWQdittK8b9AcLe"
upstream_id
"a1a480f77fb441d6b4379e2058f70bd7"
provider_responses
0
endpoint_id
"09b1fc56-3888-42bd-b7a0-926851c7f9d6"
id
"a1a480f77fb441d6b4379e2058f70bd7"
is_byok
false
latency
711
model_permaslug
"aion-labs/aion-3.5-20260923"
provider_name
"AionLabs"
status
200
total_cost
0.01571625
cache_discount
0.00229275
upstream_inference_cost
0
provider_name
"AionLabs"
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
20
adverbTagCount
2
adverbTags
0
"she said quietly [quietly]"
1
"Iyer said again [again]"
dialogueSentences
39
tagDensity
0.513
leniency
1
rawRatio
0.1
effectiveRatio
0.1
92.10%
AI-ism adverb frequency
Target: <2% AI-ism adverbs (58 tracked)
wordCount
1266
totalAiIsmAdverbs
2
found
0
adverb
"very"
count
1
1
adverb
"sharply"
count
1
highlights
0
"very"
1
"sharply"
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)
92.10%
AI-ism word frequency
Target: <2% AI-ism words (290 tracked)
wordCount
1266
totalAiIsms
2
found
0
word
"echoed"
count
1
1
word
"etched"
count
1
highlights
0
"echoed"
1
"etched"
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
79
matches
(empty)
100.00%
Filter word density
Target: ≤3% sentences with filter/hedge words
filterCount
0
hedgeCount
0
narrationSentences
79
filterMatches
(empty)
hedgeMatches
(empty)
100.00%
Gibberish response detection
Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words)
analyzedSentences
98
gibberishSentences
0
adjustedGibberishSentences
0
longSentenceCount
0
runOnParagraphCount
0
giantParagraphCount
0
wordSaladCount
0
repetitionLoopCount
0
controlTokenCount
0
repeatedSegmentCount
0
maxSentenceWordsSeen
62
ratio
0
matches
(empty)
100.00%
Markdown formatting overuse
Target: ≤5% words in markdown formatting
markdownSpans
0
markdownWords
0
totalWords
1269
ratio
0
matches
(empty)
100.00%
Missing dialogue indicators (quotation marks)
Target: ≤10% speech attributions without quotation marks
totalAttributions
13
unquotedAttributions
0
matches
(empty)
50.00%
Name drop frequency
Target: ≤1.0 per-name mentions per 100 words
totalMentions
30
wordCount
758
uniqueNames
6
maxNameDensity
1.45
worstName
"Quinn"
maxWindowNameDensity
3.5
worstWindowName
"Iyer"
discoveredNames
Harlow
1
Quinn
11
Iyer
11
Oxford
1
Eva
5
Camden
1
persons
0
"Harlow"
1
"Quinn"
2
"Iyer"
3
"Eva"
places
0
"Oxford"
globalScore
0.774
windowScore
0.5
100.00%
Narrator intent-glossing
Target: ≤2% narration sentences with intent-glossing patterns
analyzedSentences
49
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
1269
matches
(empty)
100.00%
Overuse of "that" (subordinate clause padding)
Target: ≤2% sentences with "that" clauses
thatCount
0
totalSentences
98
matches
(empty)
100.00%
Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs
44
mean
28.84
std
27.28
cv
0.946
sampleLengths
0
57
1
15
2
51
3
10
4
36
5
23
6
67
7
27
8
4
9
18
10
44
11
25
12
9
13
21
14
44
15
4
16
3
17
70
18
27
19
9
20
34
21
6
22
3
23
36
24
9
25
47
26
20
27
7
28
4
29
30
30
102
31
14
32
9
33
50
34
8
35
48
36
4
37
12
38
4
39
40
40
6
41
76
42
11
43
125
100.00%
Passive voice overuse
Target: ≤2% passive sentences
passiveCount
1
totalSentences
79
matches
0
"been drained"
91.60%
Past progressive (was/were + -ing) overuse
Target: ≤2% past progressive verbs
pastProgressiveCount
2
totalVerbs
123
matches
0
"was staring"
1
"was spinning"
100.00%
Em-dash & semicolon overuse
Target: ≤2% sentences with em-dashes/semicolons
emDashCount
1
semicolonCount
0
flaggedSentences
1
totalSentences
98
ratio
0.01
matches
0
"The needle ignored her, ignored north, ignored every direction — except one."
100.00%
Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount
758
adjectiveStacks
0
stackExamples
(empty)
adverbCount
23
adverbRatio
0.030343007915567283
lyAdverbCount
8
lyAdverbRatio
0.010554089709762533
100.00%
Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences
98
echoCount
0
echoWords
(empty)
100.00%
Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences
98
mean
12.95
std
11.72
cv
0.905
sampleLengths
0
21
1
8
2
28
3
15
4
2
5
30
6
19
7
10
8
12
9
13
10
2
11
9
12
11
13
12
14
5
15
62
16
4
17
10
18
10
19
1
20
2
21
4
22
18
23
22
24
10
25
12
26
2
27
15
28
8
29
3
30
6
31
12
32
9
33
15
34
29
35
4
36
2
37
1
38
9
39
21
40
11
41
13
42
9
43
3
44
2
45
2
46
27
47
9
48
11
49
23
90.14%
Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats
5
diversityRatio
0.5714285714285714
totalSentences
98
uniqueOpeners
56
53.76%
Adverb-first sentence starts
Target: ≥3% sentences starting with an adverb
adverbCount
1
totalSentences
62
matches
0
"Then she looked past it,"
ratio
0.016
100.00%
Pronoun-first sentence starts
Target: ≤30% sentences starting with a pronoun
pronounCount
15
totalSentences
62
matches
0
"Her torch swept the curved"
1
"She pressed the back of"
2
"She lifted the man's hand"
3
"She worked it free with"
4
"She bagged it"
5
"She walked back along them,"
6
"She crouched at the last"
7
"She had a torch strapped"
8
"she said quietly"
9
"She tucked a curl behind"
10
"She was staring at the"
11
"She crossed to it and"
12
"She sighted along it."
13
"She looked at the frost-rimed"
14
"She turned the compass over"
ratio
0.242
56.77%
Subject-first sentence starts
Target: ≤72% sentences starting with a subject
subjectCount
50
totalSentences
62
matches
0
"The ladder down into the"
1
"Detective Harlow Quinn counted them"
2
"A constable's torch beam wobbled"
3
"Her torch swept the curved"
4
"The track bed had been"
5
"The body lay in the"
6
"DS Iyer crouched beside him,"
7
"Quinn crouched opposite him."
8
"The dead man's shoes were"
9
"She pressed the back of"
10
"She lifted the man's hand"
11
"The fingers were stiff, and"
12
"She worked it free with"
13
"Symbols ringed its edge, worn"
14
"Iyer leaned in."
15
"She bagged it"
16
"Iyer stood, brushing grit from"
17
"Quinn swung her torch low"
18
"The gravel between the sleepers"
19
"The dead man's shoe prints"
ratio
0.806
0.00%
Subordinate conjunction sentence starts
Target: ≥2% sentences starting with a subordinating conjunction
subConjCount
0
totalSentences
62
matches
(empty)
ratio
0
100.00%
Technical jargon density
Target: ≤6% sentences with technical-jargon patterns
analyzedSentences
30
technicalSentenceCount
0
matches
(empty)
100.00%
Useless dialogue additions
Target: ≤5% dialogue tags with trailing filler fragments
totalTags
20
uselessAdditionCount
0
matches
(empty)
100.00%
Dialogue tag variety (said vs. fancy)
Target: ≤10% fancy dialogue tags
totalTags
9
fancyCount
0
fancyTags
(empty)
dialogueSentences
39
tagDensity
0.231
leniency
0.462
rawRatio
0
effectiveRatio
0
90.8825%