NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCSQP5QMB4E0EZYCPF7G4XJ
Run Details
Model:
Thinking Machines Inkling (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
14m
Cost:
$0.037110
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.01%
Dialogue to Total Word Ratio
(details)
0.0071%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% words of dialogue between two thieves named Molly and Bill who are planning to break into a children's party to steal all the presents. Only include the prose of the scene without any markdown or commentary.
Model response
Messages
Meta
AI
Molly adjusted the rhinestone-studded unicorn balaclava—liberated from a craft fair clearance bin—and squinted at the Henderson floor plan spread across the van’s sticky dashboard. Bill was eating cold pepperoni pizza directly over the map, leaving a translucent grease stain shaped like regret. "Bill, you're ruining the map," Molly said. "I'm adding realistic terrain," Bill replied. "Also, this pepperoni is older than democracy." The vehicle smelled of hay, poor life choices, and the chemical sweetness of stolen glitter. They had parked behind a dumpster two blocks east of the Henderson residence, where thirty sugar-crazed six-year-olds were being herded by a balloon artist named Sparkles and a magician who wore suspiciously large sleeves and eyeliner that suggested a criminal past. A rainbow arch spanned the driveway, and somewhere a child was screaming with pure, uncut joy—excellent acoustic camouflage for a felony. Three goats—borrowed for authenticity in their mobile petting zoo disguise—chewed the upholstery with philosophical indifference, occasionally bleating opinions about capitalism. Molly tapped the fogged window with a gloved finger. "We enter at noon sharp. Catering staff. Two aprons. Zero dignity." Bill swallowed, leaving crumbs in his stubble. "I can't cook. I burned water once." "You don't cook. You carry a tray of hollow cupcakes and project the disappointment of a man who watched capitalism fail." Bill adjusted his chef hat, which was actually a shower cap with a pom-pom hot-glued to the crown, and studied the party through binoculars. "We steal every present? Even the boa constrictor piñata?" "Especially the piñata. Children don't need live reptiles disguised as party favors." Bill's conscience visibly twitched beneath his beard. "These are birthday gifts, Molly. Are we monsters?" "We're professionals with flexible ethics. Remember the museum raccoon?" "That taxidermy job was emotional. This feels worse." "The parents will be mesmerized by Sparkles twisting a balloon giraffe with questionable anatomy and possibly illegal proportions. The gift table sits behind the lemonade stand, guarded only by social politeness and a parent checking a smartphone. We blend with the catering chaos, liberate the goods, and vanish in four minutes flat—before anyone realizes the cupcakes are decoys made of cardboard and regret." "I hate aprons," Bill said. "They possess no functional pockets." "That is precisely why Gertrude wears the goat satchel." Bill pointed toward the back of the van, where one goat was methodically consuming the printed escape route with alarming efficiency. "Gertrude ate our map." "Then we improvise. We always improvise." Bill sighed, finishing the crust with the resignation of a man who had made worse choices before lunch. "Fine. But if I must hum 'Happy Birthday' as auditory cover, I charge double." "You will hum like a guilty ghost who knows every lyric too well." They sat in weighted silence, listening to distant shrieks that signaled peak chaos—perfect conditions for theft. Molly pulled on her pink rhinestone gloves and checked the empty cupcake tray they had lifted from a bakery dumpster an hour earlier. "So," she said, "we steal every present and disappear before the cake is cut." "If we're caught?" "We blame the magician," Molly said. "He has guilty eyebrows and sleeves perfect for hiding guilt." Molly stood, straightening her unicorn mask until it sat with theatrical menace against her cheeks. "Let's rob some toddlers." Bill grabbed the tray, the goat leash, and a bag of decoy balloons printed with numbers that meant nothing. "For the record, this is ethically ambiguous." "Your conscience is adorable and entirely unhelpful. Now move. Sparkles is about to inflate something dangerous." The van door creaked open, releasing the smell of goat, regret, and cheap pepperoni into the warm afternoon air. Two ridiculous figures in stained aprons stepped into the sunlight, heading toward the rainbow arch and the world's most absurd felony.
created_at
"2026-07-25T14:09:15.455Z"
7/25/2026, 2:09:15 PM
model
"thinkingmachines/inkling-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
769
moderation_latency
(null)
generation_time
839918
tokens_prompt
75
tokens_completion
9258
native_tokens_prompt
75
native_tokens_completion
9151
native_tokens_completion_images
(null)
native_tokens_reasoning
8223
native_tokens_cached
32
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
0.03710999
router
(null)
provider_responses
0
endpoint_id
"d6b9d68e-115a-49d7-81cd-1ad1c6942b0e"
id
"chatcmpl-RcXmLWibsozadjHGn2pPAZqO"
is_byok
false
latency
63
model_permaslug
"thinkingmachines/inkling-20260715"
provider_name
"DeepInfra"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784988555-iMdiA9bZ7BXJkIyc4kJR"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784988555-KRQ2n8BHDdecRU4hzUWz"
upstream_id
"chatcmpl-RcXmLWibsozadjHGn2pPAZqO"
total_cost
0.03710999
cache_discount
0.00002656
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
629
0.01%
Dialogue to Total Word Ratio
Ratio: 43.31%, Deviation: 13.31%
neededClean
false
wordsTotal
635
wordsDialogue
275
0.0071%