NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD6NA77DHDF3WWWF81YTH1K
Run Details
Model:
MoonshotAI: Kimi K3 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
47.0s
Cost:
$0.018914
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
14.30%
Matches word count
(details)
48.32%
Dialogue to Total Word Ratio
(details)
31.3102%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
“Bill,” Molly whispered, crouched behind a lemonade stand shaped like a giraffe, “explain to me why our hideout has a sticker that says I’M FOUR TODAY.” “Because the Hendersons threw the box away after Lucas’s party,” Bill said. “And because you said blend in.” “I said blend into the neighborhood, not into the furniture.” “Same thing, if you’re brave.” Molly peeled a balloon off her sleeve. “The plan. Again. Slowly.” Bill unrolled a crayon map across the curb. “At two o’clock, twenty children arrive for Princess Dinosaur Unicorn Adventure. At two-oh-five, the mom puts the presents on the picnic table. At two-ten, the clown starts juggling, which creates a natural distraction.” “A clown is not natural.” “In suburbia he is.” “Continue.” “We wear these.” Bill held up two costumes: one kangaroo, one mailman. “Why am I the kangaroo?” “Pockets.” “Fair. Why are you the mailman on a Sunday?” “Authority.” “You get hives near authority.” “Only elected authority.” Molly tapped the map. “And the presents?” “We replace them with potatoes wrapped in shiny paper.” “Bill.” “It teaches patience.” “It teaches rage.” “Fine, onions for the older kids.” Molly shut her eyes. “Remind me why we’re stealing from children.” “Because the Big Score needs seed money, because Vernon sells stolen kazoos to toddlers, and because the Henderson boy got a solid-gold yo-yo while I ate cereal for dinner.” “You’re forty-one.” “Cereal has no age limit.” A sprinkler coughed on across the hedge. Both of them flattened into the grass like guilty rugs. “New problem,” Bill said. “The dad is grilling.” “Men grill at funerals. It means nothing.” “He’s wearing an apron that says GRILL SERGEANT.” “That means he’s insecure.” “He’s holding tongs like nunchucks.” “Then we avoid tong range.” “What’s tong range?” “Four feet, or one bad opinion about barbecue sauce.” Molly crawled to the fence and peeked through a knothole. “I see the gift table. There’s a box big enough to hide a baby grand.” “Or a baby.” “Don’t say baby during a heist.” “Sorry. A small piano.” “Better.” Bill checked a watch with no hands. “We go when the cake comes out.” “Why?” “Fire, singing, crying. Nature’s smoke bomb.” “That’s a birthday party.” “Exactly. Organized panic with frosting.” Molly sighed. “If a child sees us, what do we do?” “Say we’re gift inspectors.” “Gift inspection is not a thing.” “It is if you carry a clipboard.” “Do you have a clipboard?” Bill produced one from the kangaroo pouch. “That was my pouch.” “Shared pouch, shared destiny.” A chorus of small voices shrieked inside the house. “They’ve started musical chairs,” Molly said. “Violent game. Chairs move, alliances break.” “Focus.” “I am focused. I’m also terrified.” “Good. Terror is cardio.” Bill raised the mailman cap. “For rent money?” “For rent money,” Molly said. “For Vernon’s stupid kazoo empire?” “Mostly against it.” “For potatoes?” “Don’t push it.” They climbed over the fence into a wall of noise, glitter, and tiny shoes. A little girl pointed at Bill and screamed, “The mail is here!” Bill froze. “Cover’s blown.” Molly grabbed his sleeve. “Then deliver something, mailman.” “What?” “Chaos.” “Finally,” Bill said, “a job I understand.”
created_at
"2026-07-25T17:55:09.164Z"
7/25/2026, 5:55:09 PM
model
"moonshotai/kimi-k3-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2352
moderation_latency
(null)
generation_time
46988
tokens_prompt
75
tokens_completion
1429
native_tokens_prompt
147
native_tokens_completion
1258
native_tokens_completion_images
(null)
native_tokens_reasoning
387
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
0.0189141
router
(null)
provider_responses
0
endpoint_id
"650c9b12-7b6d-444d-8d92-df8daef84d7a"
id
"chatcmpl-6a64f869e53def91dc7e7814"
is_byok
false
latency
2352
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-1785002109-gY8NnZmJIYdGIES7sZ68"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785002109-0KNESq9UPMSo512o5kq2"
upstream_id
"chatcmpl-6a64f869e53def91dc7e7814"
total_cost
0.0189141
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
14.30%
Matches word count
n/a
neededClean
false
words
521
48.32%
Dialogue to Total Word Ratio
Ratio: 73.81%, Deviation: 3.81%
neededClean
false
wordsTotal
527
wordsDialogue
389
31.3102%