NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCWB6KB99NG5HM86XJ4CX04
Run Details
Model:
Thinking Machines Inkling (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
21.2s
Cost:
$0.012986
Provider:
BaseTen
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 unfurled the crumpled blueprint over a pizza box stained with yesterday’s pepperoni grease. The paper was covered in her neat handwriting and one suspicious coffee ring that looked like a target. “Listen up. The Henderson spawn hits seven years old tomorrow. Party kicks off at two sharp. By three, the kids are in a sugar coma, the parents are three boxes of Pinot deep, and the gift table is essentially a small retail warehouse.” Bill squinted at a crayon diagram labeled in glitter glue: Securtie Plan Phase One. “You want us to rob a room full of first-graders?” “We’re not robbing children, Bill. We are liberating merchandise from negligent storage practices. There’s a PlayStation Five on that registry. I verified.” Bill scratched his stubble. “There’s also a balloon dinosaur and, according to this, an actual guy in a dinosaur costume. That’s redundant. Too many reptiles.” “We neutralize the dinosaur,” Molly said, jabbing a red circle with her pen. “You run distraction protocols—” “I’m not licensed for magic.” “You carry a deck of cards with naked ladies printed on them. That counts as sleight of hand in a crisis.” Bill exhaled through his nose. “What about the adults? They notice things.” Molly slid a laminated schedule across the box. “Mom does the wine. Dad tends a propane death-trap called a grill, then mentally checks out. Grandparents roll in at four with more poorly wrapped boxes, stay for exactly one slice of sheet cake, then flee to play bridge. Our window is two-fifteen to two-forty-five.” “Thirty minutes? For a heist?” “For professionals? Practically a vacation.” Bill leaned back in the folding chair, which groaned like a wounded moose. “Last time you said that, I woke up in a dumpster arguing with a raccoon.” “That raccoon was an asset. He had excellent night vision and zero tax liability.” “Right. What’s our ingress?” “Service door by the bouncy castle. The children are contained inside an inflatable prison. We slip in, bag the goods, egress through the garage while everyone is singing that one awful song about a shark.” “And if some kid spots us?” Molly’s grin turned sharp. “Then we become the mysterious ‘fun surprise grown-ups.’ I’ve packed a burlap sack and a synthetic beard.” “That is Santa Claus. It’s mid-July.” “Children do not understand seasonal logistics. They understand presents leaving with strangers.” Bill studied the map with the intensity of a man deciphering ancient runes. “Stash location?” “Panel van across the residential street. Engine running, radio on low. You drive. The brakes are questionable, but the anonymity is excellent.” “I always drive. Last time—” “Because you are unparalleled at parallel parking under pressure, and I am unparalleled at not leaving fingerprints on security cameras.” She snapped the blueprint shut. “Dress code is black. Move silent. Do not engage the children in conversation. Do not accept juice boxes. Do not, under any circumstance, enter the bouncy castle. It is structurally unsound and emotionally dangerous.” “Can I talk to the cake?” “Only if it is chocolate. Otherwise, maintain radio silence.” Bill rose, stretching his back with a series of cracks that sounded like microwave popcorn. “This is officially the strangest gig we’ve pulled.” “That’s only because you’ve never witnessed forty gifts accumulated beneath a single plastic tablecloth with cartoon ponies on it.” Molly zipped her windbreaker and checked her watch. “Tomorrow we hit at sugar o’clock. Be punctual.” Bill grabbed his baseball cap. “If that dinosaur entertainer actually exists, I am filing a complaint with management.” “He exists. His name is Greg. He is terrified of emotional intimacy, according to the party planner’s public social media.” “How do you even—” “I conduct research, Bill. It’s called professionalism.” Molly opened the door, letting in a breeze that smelled like suburban despair. “Noon sharp. Wear sneakers that don’t squeak. And leave the naked cards at home; this is a family establishment, technically.” Bill paused in the doorway. “You’re enjoying this too much.” “I’m a thief with a calendar. Of course I’m enjoying it.” She flicked the lights off. “See you at the drop point. Try not to get adopted.” Bill adjusted his cap. “If I’m not back by midnight, send the raccoon.” “He’s retired. Try not to need rescue.” She stepped into the hallway. “Tomorrow, Bill. We strike at cake time.”
created_at
"2026-07-25T14:54:52.019Z"
7/25/2026, 2:54:52 PM
model
"thinkingmachines/inkling-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
128
moderation_latency
(null)
generation_time
19615
tokens_prompt
75
tokens_completion
3716
native_tokens_prompt
75
native_tokens_completion
3188
native_tokens_completion_images
(null)
native_tokens_reasoning
2223
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
0.0129864
router
(null)
provider_responses
0
endpoint_id
"ba2df08c-c0a0-4e0e-9f31-d79a9c475b70"
id
"chatcmpl-1028d8a222884b158bd13ebdb66c9022"
is_byok
false
latency
128
model_permaslug
"thinkingmachines/inkling-20260715"
provider_name
"BaseTen"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784991292-F7SKvbLmhNtgefy0tUOC"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784991292-aykNWln9Qx0atN0KgUHe"
upstream_id
"chatcmpl-1028d8a222884b158bd13ebdb66c9022"
total_cost
0.0129864
cache_discount
(null)
upstream_inference_cost
0
provider_name
"BaseTen"
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
715
0.00%
Dialogue to Total Word Ratio
Ratio: 76.87%, Deviation: 46.87%
neededClean
false
wordsTotal
722
wordsDialogue
555
0.0000%