NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD4S1FSR18TCNREJ05WGEHZ
Run Details
Model:
MoonshotAI: Kimi K3 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
51.1s
Cost:
$0.019649
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
3.62%
Matches word count
(details)
18.08%
Dialogue to Total Word Ratio
(details)
10.8498%
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
The van smelled of cheese puffs and poor decisions. Molly spread their map across the steering wheel, though it was mostly a napkin with a drawing of a house, a cake, and several arrows pointing at a stick figure labeled BIRTHDAY GREMLIN. Bill squinted. “Is the gremlin hostile?” “That’s Timmy,” Molly said. “Age six. Likes dinosaurs. Hates bedtime.” “Then we negotiate.” “We are not negotiating with a six-year-old.” “Molly, every empire falls because somebody underestimated a six-year-old.” She tapped the napkin. Outside, the party thumped with off-key singing and the shriek of balloons being murdered by chairs. Through the window they could see paper hats bobbing like a fleet of tiny doomed ships. “Plan,” Molly said. “We slip in during musical statues. Everybody freezes. We don’t. We take presents. We leave.” Bill nodded solemnly. “Flawless. What if somebody looks at us?” “Freeze harder.” “Right. Emotionally or physically?” “Both.” He peeled the wrapper from a cheese puff with the delicacy of a safecracker. “And the clown?” Molly went still. “What clown?” A honk came from the garden. Then another, closer, as if the first honk had called for backup. Bill slowly turned the rear-view mirror. A face painted like a deranged sunrise appeared in it. Molly whispered, “Drive.” “We’re parked.” “Then parked menacingly.” The clown knocked on the window. Bill rolled it down two inches, which somehow made the clown bigger. “Oh no,” said the clown. “Oh yes,” said Bill. Molly leaned over. “We’re from the Present Inspection Bureau.” The clown frowned. “Is that real?” “It’s new,” Molly said. “Very new,” Bill added. “Still wet.” The clown’s nose squeaked as he considered this. Somewhere inside, a child screamed, “PASS THE PARCEL,” and forty tiny feet stampeded like wildebeest in party shoes. Molly gripped the dashboard. “Bill, the presents are in the lounge.” “Then we lounge adjacent.” “We go in, compliment the cake, locate the gifts, create a diversion, and exit with dignity.” “What sort of diversion?” “You juggle.” “I can’t juggle.” “Perfect. Nobody suspects failure.” The clown opened the van door. “You lot coming in? Only it’s jelly time.” Molly and Bill looked at each other. Years of crime had prepared them for lasers, dogs, lasers on dogs, and one memorable incident involving a haunted duvet. They had not prepared for jelly. Bill cleared his throat. “We have a warrant.” “For jelly?” “Surprise jelly warrant.” Inside, the table sagged under wrapped boxes: dinosaurs, dolls, a science kit already leaking glitter, a drum set somebody would regret by Tuesday. Molly’s hand hovered over a silver package shaped like greed. Then Timmy appeared, solemn as a judge, holding out a paper plate. “You can have cake,” he said, “if you didn’t come to steal.” Molly froze. Bill froze harder, emotionally and physically. “Us?” Molly said. “Never.” “Good,” said Timmy. “Because Aunt Carol bites.” From the kitchen came a growl that was almost certainly Aunt Carol. Bill took the cake. “New plan.” Molly sighed, already reaching for napkins. “We steal nothing, eat quickly, and leave before musical chairs.” “And the presents?” Timmy pointed at Aunt Carol, who was guarding them with a spoon. Bill nodded. “Retirement sounds nice.”
created_at
"2026-07-25T17:22:14.144Z"
7/25/2026, 5:22:14 PM
model
"moonshotai/kimi-k3-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2482
moderation_latency
(null)
generation_time
51042
tokens_prompt
75
tokens_completion
1490
native_tokens_prompt
147
native_tokens_completion
1307
native_tokens_completion_images
(null)
native_tokens_reasoning
486
native_tokens_cached
147
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
0.0196491
router
(null)
provider_responses
0
endpoint_id
"650c9b12-7b6d-444d-8d92-df8daef84d7a"
id
"chatcmpl-6a64f0c7689654d454e9fac7"
is_byok
false
latency
2482
model_permaslug
"moonshotai/kimi-k3-20260715"
provider_name
"Moonshot AI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1785000134-1k67uUzR3L00VhWgs02e"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785000134-aCOsaIrMkQfHkDfAhSbc"
upstream_id
"chatcmpl-6a64f0c7689654d454e9fac7"
total_cost
0.0196491
cache_discount
0.0003969
upstream_inference_cost
0
provider_name
"Moonshot AI"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
3.62%
Matches word count
n/a
neededClean
false
words
524
18.08%
Dialogue to Total Word Ratio
Ratio: 35.85%, Deviation: 5.85%
neededClean
false
wordsTotal
530
wordsDialogue
190
10.8498%