Run: 2026-02-19-01KHTYKJWZS07RX0FXPEPAK2H8

Run Details

Model:
DeepSeek V3 (2024-12-26)
Test:
Bad Writing Habits
Scenario:
Mystery: examining a crime scene
Duration:
23.5s
Cost:
$0.001403
Provider:
Novita

Evaluation summary

Score Evaluator
Details Link
100.00% Adverbs in dialogue tags (details)
80.62% AI-ism adverb frequency (details)
100.00% AI-ism character names (details)
100.00% AI-ism location names (details)
32.17% 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)
60.18% Name drop frequency (details)
45.83% 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)
97.47% Passive voice overuse (details)
100.00% Past progressive (was/were + -ing) overuse (details)
0.00% Em-dash & semicolon overuse (details)
93.82% Purple prose (modifier overload) (details)
100.00% Repeated phrase echo (details)
100.00% Sentence length variance (details)
68.89% Sentence opener variety (details)
79.37% Adverb-first sentence starts (details)
100.00% Pronoun-first sentence starts (details)
55.24% Subject-first sentence starts (details)
0.00% Subordinate conjunction sentence starts (details)
100.00% Technical jargon density (details)
25.00% Useless dialogue additions (details)
100.00% Dialogue tag variety (said vs. fancy) (details)
81.2862% 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

AI
The abandoned Tube station beneath Camden reeked of damp concrete and something metallic—blood, maybe, or rust. Detective Harlow Quinn ducked under the police tape, her boots crunching on broken glass. The emergency lights flickered, casting long shadows across the cracked tiles. "Quinn." A uniformed officer nodded at her, his face tight. "You're gonna want to see this." She followed him deeper into the station, past boarded-up ticket booths and graffiti-strewn pillars. The air grew colder. At the edge of the platform, a body lay sprawled across the tracks, limbs twisted at unnatural angles. The man’s face was pale, lips tinged blue, but there wasn’t a drop of blood on him. "Who found him?" "Market trader," the officer said. "Said the guy just... dropped. No scream, no struggle." Quinn crouched, examining the victim’s hands. No defensive wounds. No signs of a fight. His coat was expensive, tailored, but his pockets were empty—no wallet, no phone. Just a small brass compass tucked into his inner pocket. She pried it open. The needle spun lazily, then jerked to the left, toward the tunnel’s black mouth. "Morris would’ve called this a 'clean kill,'" a voice said behind her. Quinn didn’t turn. "Morris isn’t here." Detective Simmons stepped into the dim light, hands in his pockets. "No signs of trauma, no poison in the prelim tox screen. Heart just stopped." He nudged the compass with his shoe. "Probably fell off some tourist." She snapped the compass shut. "Tourists don’t carry antiques with sigils etched into them." Simmons sighed. "You always do this. Look for the weird angle. Maybe the guy had a heart condition. Maybe he owed the wrong people money." Quinn stood, rolling the compass between her fingers. The casing was warm. "Then where’s his wallet? His phone? This wasn’t a mugging—no bruising, no signs of force. And his shoes are spotless. You think he walked down here in the dark, in Oxford brogues, without scuffing them once?" Simmons opened his mouth, then shut it. She turned toward the tunnel. The compass needle twitched again. "This place is part of the Veil Market. You need a token to get in. So how’d our victim get here?" "Maybe he had one." "Then where is it?" Simmons rubbed his temple. "You’re saying someone lured him here, took his token, and... what? Scared him to death?" Quinn’s watch ticked loudly in the silence. She scanned the platform again—the scuff marks near the edge, the way the dust was disturbed in a perfect circle around the body. No footprints leading away. "Not scared," she said slowly. "Taken." Simmons scoffed. "Taken by what, ghosts?" She didn’t answer. The compass hummed faintly in her palm. The needle pointed straight at the tunnel. And then, for just a second, the lights went out. When they flickered back on, the compass was ice-cold. The needle spun wildly. Simmons hadn’t noticed. He was already walking away, radio crackling. "Wrap it up, Quinn. Dead end." She stared into the dark tunnel. The air smelled like ozone now. Not a dead end. A door.
created_at"2026-02-19T12:39:42.246Z" 2/19/2026, 12:39:42 PM
model"deepseek/deepseek-chat-v3"
app_id182717
external_user(null)
streamedtrue
cancelledfalse
latency1459
moderation_latency(null)
generation_time23472
tokens_prompt1149
tokens_completion828
native_tokens_prompt1032
native_tokens_completion762
native_tokens_completion_images0
native_tokens_reasoning0
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"
usage0.0014034
router(null)
provider_responses
0
endpoint_id"cefb5a1e-9fea-4496-9a64-4e1b11a7cc8c"
id"9372170a1056e0c183e0eb9eb3750b03"
is_byokfalse
latency1459
model_permaslug"deepseek/deepseek-chat-v3"
provider_name"Novita"
status200
api_type"completions"
id"gen-1771504782-EU8zxxGSOzwgHVzDIhzE"
upstream_id"9372170a1056e0c183e0eb9eb3750b03"
total_cost0.0014034
cache_discount(null)
upstream_inference_cost0
provider_name"Novita"

Evaluation details

Result Evaluator Details Meta Data
100.00% Adverbs in dialogue tags
Target: ≤10% dialogue tags with adverbs
totalTags5
adverbTagCount1
adverbTags
0"she said slowly [slowly]"
dialogueSentences20
tagDensity0.25
leniency0.5
rawRatio0.2
effectiveRatio0.1
80.62% AI-ism adverb frequency
Target: <2% AI-ism adverbs (58 tracked)
wordCount516
totalAiIsmAdverbs2
found
0
adverb"lazily"
count1
1
adverb"slowly"
count1
highlights
0"lazily"
1"slowly"
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)
32.17% AI-ism word frequency
Target: <2% AI-ism words (290 tracked)
wordCount516
totalAiIsms7
found
0
word"flickered"
count2
1
word"tinged"
count1
2
word"etched"
count1
3
word"silence"
count1
4
word"scanned"
count1
5
word"perfect"
count1
highlights
0"flickered"
1"tinged"
2"etched"
3"silence"
4"scanned"
5"perfect"
100.00% Cliché density
Target: ≤1 cliche(s) per 800-word window
totalCliches0
maxInWindow0
found(empty)
highlights(empty)
100.00% Emotion telling (show vs. tell)
Target: ≤3% sentences with emotion telling
emotionTells0
narrationSentences45
matches(empty)
100.00% Filter word density
Target: ≤3% sentences with filter/hedge words
filterCount1
hedgeCount0
narrationSentences45
filterMatches
0"watch"
hedgeMatches(empty)
100.00% Gibberish response detection
Target: ≤1% gibberish-like sentences (hard fail if a sentence exceeds 800 words)
analyzedSentences60
gibberishSentences0
adjustedGibberishSentences0
longSentenceCount0
runOnParagraphCount0
giantParagraphCount0
wordSaladCount0
repetitionLoopCount0
controlTokenCount0
maxSentenceWordsSeen37
ratio0
matches(empty)
100.00% Markdown formatting overuse
Target: ≤5% words in markdown formatting
markdownSpans0
markdownWords0
totalWords509
ratio0
matches(empty)
100.00% Missing dialogue indicators (quotation marks)
Target: ≤10% speech attributions without quotation marks
totalAttributions6
unquotedAttributions0
matches(empty)
60.18% Name drop frequency
Target: ≤1.0 per-name mentions per 100 words
totalMentions14
wordCount334
uniqueNames5
maxNameDensity1.8
worstName"Simmons"
maxWindowNameDensity3
worstWindowName"Simmons"
discoveredNames
Tube1
Camden1
Harlow1
Quinn5
Simmons6
persons
0"Camden"
1"Harlow"
2"Quinn"
3"Simmons"
places(empty)
globalScore0.602
windowScore0.667
45.83% Narrator intent-glossing
Target: ≤2% narration sentences with intent-glossing patterns
analyzedSentences24
glossingSentenceCount1
matches
0"smelled like ozone now"
100.00% "Not X but Y" pattern overuse
Target: ≤1 "not X but Y" per 1000 words
totalMatches0
per1kWords0
wordCount509
matches(empty)
100.00% Overuse of "that" (subordinate clause padding)
Target: ≤2% sentences with "that" clauses
thatCount0
totalSentences60
matches(empty)
100.00% Paragraph length variance
Target: CV ≥0.5 for paragraph word counts
totalParagraphs26
mean19.58
std15.58
cv0.796
sampleLengths
041
116
253
33
414
555
612
76
837
914
1025
1148
127
1331
144
154
1619
1734
186
196
2017
2110
2213
2316
2412
256
97.47% Passive voice overuse
Target: ≤2% passive sentences
passiveCount1
totalSentences45
matches
0"was disturbed"
100.00% Past progressive (was/were + -ing) overuse
Target: ≤2% past progressive verbs
pastProgressiveCount0
totalVerbs61
matches(empty)
0.00% Em-dash & semicolon overuse
Target: ≤2% sentences with em-dashes/semicolons
emDashCount3
semicolonCount0
flaggedSentences3
totalSentences60
ratio0.05
matches
0"The abandoned Tube station beneath Camden reeked of damp concrete and something metallic—blood, maybe, or rust."
1"His coat was expensive, tailored, but his pockets were empty—no wallet, no phone."
2"She scanned the platform again—the scuff marks near the edge, the way the dust was disturbed in a perfect circle around the body."
93.82% Purple prose (modifier overload)
Target: <4% adverbs, <2% -ly adverbs, no adj stacking
wordCount340
adjectiveStacks0
stackExamples(empty)
adverbCount16
adverbRatio0.047058823529411764
lyAdverbCount5
lyAdverbRatio0.014705882352941176
100.00% Repeated phrase echo
Target: ≤20% sentences with echoes (window: 2)
totalSentences60
echoCount0
echoWords(empty)
100.00% Sentence length variance
Target: CV ≥0.4 for sentence word counts
totalSentences60
mean8.48
std6.57
cv0.774
sampleLengths
016
114
211
310
46
514
64
718
817
93
105
119
126
133
145
1513
1610
174
1814
1912
203
213
2211
2321
245
255
269
272
2823
298
304
3136
327
335
345
3521
364
374
384
3915
407
4123
424
435
441
452
464
473
487
497
68.89% Sentence opener variety
Target: ≥60% unique sentence openers
consecutiveRepeats2
diversityRatio0.43333333333333335
totalSentences60
uniqueOpeners26
79.37% Adverb-first sentence starts
Target: ≥3% sentences starting with an adverb
adverbCount1
totalSentences42
matches
0"Just a small brass compass"
ratio0.024
100.00% Pronoun-first sentence starts
Target: ≤30% sentences starting with a pronoun
pronounCount11
totalSentences42
matches
0"She followed him deeper into"
1"His coat was expensive, tailored,"
2"She pried it open."
3"He nudged the compass with"
4"She snapped the compass shut."
5"She turned toward the tunnel."
6"She scanned the platform again—the"
7"she said slowly"
8"She didn’t answer."
9"He was already walking away,"
10"She stared into the dark"
ratio0.262
55.24% Subject-first sentence starts
Target: ≤72% sentences starting with a subject
subjectCount34
totalSentences42
matches
0"The abandoned Tube station beneath"
1"Detective Harlow Quinn ducked under"
2"The emergency lights flickered, casting"
3"A uniformed officer nodded at"
4"She followed him deeper into"
5"The air grew colder."
6"The man’s face was pale,"
7"the officer said"
8"Quinn crouched, examining the victim’s"
9"His coat was expensive, tailored,"
10"She pried it open."
11"The needle spun lazily, then"
12"a voice said behind her"
13"Quinn didn’t turn."
14"Detective Simmons stepped into the"
15"He nudged the compass with"
16"She snapped the compass shut."
17"Quinn stood, rolling the compass"
18"The casing was warm."
19"Simmons opened his mouth, then"
ratio0.81
0.00% Subordinate conjunction sentence starts
Target: ≥2% sentences starting with a subordinating conjunction
subConjCount0
totalSentences42
matches(empty)
ratio0
100.00% Technical jargon density
Target: ≤6% sentences with technical-jargon patterns
analyzedSentences13
technicalSentenceCount0
matches(empty)
25.00% Useless dialogue additions
Target: ≤5% dialogue tags with trailing filler fragments
totalTags5
uselessAdditionCount1
matches
0"officer nodded, his face tight"
100.00% Dialogue tag variety (said vs. fancy)
Target: ≤10% fancy dialogue tags
totalTags3
fancyCount0
fancyTags(empty)
dialogueSentences20
tagDensity0.15
leniency0.3
rawRatio0
effectiveRatio0
81.2862%