NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCX0YCFYG62RCQZFFEVE29G
Run Details
Model:
Thinking Machines Inkling (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 49s
Cost:
$0.052109
Provider:
Together
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
89.88%
Dialogue to Total Word Ratio
(details)
44.9407%
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 flattened the creased blueprint across a folding table that listed like a drunken sailor, anchoring the corner with a burrito that smelled of regret, forgotten salsa, and questionable meat origins. The basement reeked of mildew, expired dreams, and the sharp chemical scent of stolen printer ink. A single bulb dangled overhead, flickering with the enthusiasm of a dying firefly, painting Molly's gold tooth in stuttering gold and deep shadow. She wore a leather jacket scarred by three continents and one unfortunate zoo fence. Her crimson nails traced slow, predatory circles around a crudely drawn bouncy castle marked with a bright red X. Bill hovered behind her, adjusting mirrored sunglasses with the solemnity of a man preparing to identify a body—specifically, his own career. His beard had achieved the texture of abandoned wool left in a storm drain. "Twelve presents?" "Confirmed," Molly said, tapping the paper with a nail that could open cans. "Motorized unicorn, four hundred retail. Parents are distracted tech executives who genuinely believe parenting is a notification they can swipe away." Bill scratched through his jaw, leaving lines in the grime. "We risk prison for plastic?" "We redistribute excess with balloons and poor taste." Bill dragged over a chair that screamed in protest and collapsed into it with theatrical defeat. He pulled open the garment bag Molly had tossed onto the concrete like a dead body. Inside hung two suits of whiteface horror—baggy sleeves, oversized shoes, and one particularly disturbing ruffled collar that suggested a carnival nightmare. "Clown suits?" "Mournful clowns," Molly said, adjusting her collar. "Sad performers never get questioned. You juggle decoys. I bag assets." Bill straightened a polka-dot tie that waged open war against suspenders older than his criminal record. His shoes, stolen from a bowling alley in 2017, clicked against the floor. "There's always security. Some uncle named Gary who watched too many action movies." "The magician begins at four thirty," Molly said, marking an X near a fluffy doodle of what might have been a poodle. "Every adult will be mesmerized by questionable hygiene and a rabbit with obvious grievances. Seven-minute window." Bill pointed at a sketch of a fluffy nightmare near the entry point. "Canine?" "Peanut butter treats. Non-lethal, mostly legal." "Seven minutes?" "Out before cake." Bill leaned back; the chair groaned like a dying whale. "If a child addresses me?" "Say the unicorn ran away. Tell them you're emotionally unavailable." "I practice that every morning." Molly folded the map with surgical precision and slid it into her jacket pocket beside a folded switchblade that had never seen a legal day. "We arrive at three forty-five. Van running cold, sacks loaded in the trunk. If anyone stops us, we're entertainment for twins that never existed." Bill stood, adjusting cuffs that had seen better decades and worse laundering. "This plan has more holes than my socks." "Wear shoes. Don't consume the bait." Bill patted his coat pocket where a glass jar clinked softly against loose change and a single button. "Molly, are we actually stealing from children?" "We take excess. Those children have abundance and zero work ethic. We are economic heroes in greasepaint." Bill grinned, revealing a chipped canine tooth that caught the bulb's final flicker. "Best job." Molly grabbed her duffel—black canvas stained with what might have been coffee, old conspiracies, or both—and headed for the stairs. The bulb finally surrendered, plunging the basement into velvet darkness for one theatrical heartbeat before gray afternoon light spilled through the door above. Bill followed, stepping into the glow with the cautious optimism of a man who had forgotten where he parked both his van and his soul. "Dignity stays?" "Next to your conscience. In the jar." "Perfect."
created_at
"2026-07-25T15:06:44.506Z"
7/25/2026, 3:06:44 PM
model
"thinkingmachines/inkling-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
160
moderation_latency
(null)
generation_time
168857
tokens_prompt
75
tokens_completion
12715
native_tokens_prompt
75
native_tokens_completion
12848
native_tokens_completion_images
(null)
native_tokens_reasoning
12041
native_tokens_cached
0
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
0.0521094
router
(null)
provider_responses
0
endpoint_id
"9911deb9-f44e-4e18-87fe-9d7b60acee72"
id
"otYdbob-6z2FuN-a20c11fccb7b4c92"
is_byok
false
latency
160
model_permaslug
"thinkingmachines/inkling-20260715"
provider_name
"Together"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784992004-DTua9onNVabP3JXvfywl"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784992004-xhygTp2qtWdJAqf7uDPc"
upstream_id
"otYdbob-6z2FuN-a20c11fccb7b4c92"
total_cost
0.0521094
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Together"
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
619
89.88%
Dialogue to Total Word Ratio
Ratio: 31.46%, Deviation: 1.46%
neededClean
false
wordsTotal
623
wordsDialogue
196
44.9407%