NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-27-01M5YF750FHX7NFET9G0GC5XVJ
Run Details
Model:
Kimi K3.1 Pro (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
4m 33s
Cost:
$0.096142
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
85.04%
Dialogue to Total Word Ratio
(details)
92.5221%
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
Molly unrolled the blueprint across the bonnet of Bill's car. It was a crayon picture of a castle, stolen from the party invitation. “Right,” Bill said. “I go in through the window.” “That’s a bouncy castle.” “I go in bouncy, then.” “You’ll bounce out again with a present stuck to your face.” “Not if I wear the false moustache.” Molly looked at him. “The moustache isn’t a disguise. It’s a cry for help.” Bill folded his arms. “Fine, mastermind. What’s your plan?” “We enter as entertainers. You’re a clown. I’m the astonishing Madame Gizmo.” “What do I do?” “Nothing. For once, your natural talents are useful.” “And what do you do?” “Distract the children with magic while you move the presents into our cake.” Bill stared at the enormous cardboard cake beside the car. “Why is there a cake?” “Because no one suspects cake.” “I suspect that cake. It has wheels.” “Custom cake.” “It has an exhaust pipe.” “Luxury custom cake.” A kazoo squeaked inside Bill’s pocket. He took it out and blew a solemn note. “Is that our getaway signal?” Molly asked. “No. This is.” He blew again, producing a noise like a goose entering a tax meeting. “Perfect. Every parent will look at you.” “Thank you.” “That wasn’t praise.” Bill consulted the invitation. “Party starts at two. We’ve got fourteen children, one magician, and a woman called Linda who has threatened to bring her own karaoke machine.” “We can use Linda.” “How?” “At exactly two fifteen, Linda sings. At two sixteen, every window in the street shatters. We stroll in wearing earplugs and expressions of concern.” “And the presents?” “Under the table by the jelly.” “Can I have some jelly?” “No.” “One spoonful?” “No.” “Gigantic spoonful?” Molly tapped the drawing. “Focus. We load the presents, start the cake, and meet you behind the shed.” “What if the shed’s locked?” “It’s a shed, not Fort Knox.” “What if there’s a dog?” “Bring a sausage.” “What if the dog’s vegetarian?” “Bring a disappointing sausage.” Bill nodded gravely and wrote this down. “And if we’re caught?” “You do the clown routine.” “I don’t have a clown routine.” “Exactly. They’ll beg us to leave.” “What about party bags?” “We’re not stealing party bags.” “But they contain bubbles.” “You own bubbles.” “These are criminal bubbles.” “Bill.” “And possibly a kazoo.” “You already have a kazoo.” “That is a terrible attitude, Molly.” “I need a backup. What if my first kazoo is injured during the operation?” “Then I can finish the job in blissful, kazooless silence.” Molly checked her watch. Somewhere beyond the hedge, children cheered as the bouncy castle inflated with a groan. “Ready?” she said. Bill put on his red nose. “Born ready.” “You were born wearing that nose?” “My mother had a difficult time.” They lifted the cake. It immediately rolled downhill, through the hedge, and into the party. From inside came Linda’s delighted cry: “Ooh! A karaoke cake!” Bill turned to Molly. “New plan?” Molly sighed. “Yes. We steal Linda.”
created_at
"2026-10-27T10:41:40.376Z"
10/27/2026, 10:41:40 AM
model
"moonshotai/kimi-k3.1-pro-20261023"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1901
moderation_latency
(null)
generation_time
273074
tokens_prompt
75
tokens_completion
8794
native_tokens_prompt
146
native_tokens_completion
9585
native_tokens_completion_images
(null)
native_tokens_reasoning
8783
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
0.096142
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.1; linux; x64))"
http_referer
(null)
request_id
"req-1793097700-XDYHLp4Lfn4EZTaEAI4h"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1793097700-6ZPVFDUJTKFpr7ge4q3Z"
upstream_id
"chatcmpl-6ae07fe545ef637c96a5d9b4"
provider_responses
0
endpoint_id
"a3fafaac-f097-4ef1-86b3-842854dad5c6"
id
"chatcmpl-6ae07fe545ef637c96a5d9b4"
is_byok
false
latency
1901
model_permaslug
"moonshotai/kimi-k3.1-pro-20261023"
provider_name
"Moonshot AI"
status
200
total_cost
0.096142
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Moonshot AI"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
85.04%
Dialogue to Total Word Ratio
Ratio: 71.80%, Deviation: 1.80%
neededClean
false
wordsTotal
500
wordsDialogue
359
92.5221%