NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-27-01M5YF5E6ZXM0ZA8THNPMR6EYQ
Run Details
Model:
Kimi K3.1 Pro (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
4m 2s
Cost:
$0.080222
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
71.32%
Dialogue to Total Word Ratio
(details)
85.6598%
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 spread the birthday invitation across Bill’s kitchen table, weighting its corners with a mug, a wrench, and half a doughnut. Balloons printed around the border seemed to watch them. Bill moved the doughnut before it could compromise the operation. “Tomorrow at two,” she said. “Twenty guests, one magician, and a present table beside the conservatory. We go in as caterers, fill three cake boxes, and leave before anyone notices the jelly.” Bill leaned closer, frowning at a drawing of a smiling dinosaur. “Why would anyone notice the jelly?” Bill asked. “Because you’ll be carrying it,” Molly said. On the invitation, Mrs. Pembroke promised games, tea, and an appearance by a real fire engine. Molly had underlined fire engine twice. Bill had drawn a moustache on the dinosaur. Their previous job, at a bankers’ awards dinner, had failed because Bill tried to improve the seating plan. This time, Molly intended to control every variable, including Bill’s shoes, which currently did not match. He straightened. The frown vanished, replaced by the solemn expression he usually reserved for menus and unattended luggage. “I could be the magician. I already own a cape, and children are famously easy to impress. Last week your nephew applauded me for finding my own shoes,” Bill said. “You were wearing them on your hands. Besides, the magician is called Professor Marvo, and he has a rabbit,” Molly said. “I have a rabbit. It’s a hat with ears, but the principle is sound,” Bill replied. Molly tapped the conservatory on her sketch. It was a remarkably detailed sketch, considering she had drawn it on the back of a takeaway menu. A small cross marked the present table. Another cross marked the kitchen. A third, much larger cross marked Bill. “No magic. No juggling. No telling the birthday boy that his bicycle has been selected for a government experiment,” she said. “You never let me develop my material,” Bill said. From the hallway came a soft tearing sound. Bill’s terrier, Cromwell, was eating the second catering apron. Molly rescued it, brushed off a button, and wrote lint roller beneath Catastrophe. Bill looked offended when she added spare trousers. Cromwell swallowed the button with the discretion of a trained accomplice. She turned the menu over and revealed three lists headed Entry, Exit, and Catastrophe. The last was already the longest. “We enter, touch nothing until the cake arrives, and never, ever improvise,” Molly said. “So my cape with the hidden cake boxes is out?” Bill asked. Molly studied him across the table. Somewhere beyond the window, an ice cream van played three notes and died. Bill waited, one hand resting protectively on the cape he had somehow produced from beneath his chair. At last she folded the invitation, slid it into her pocket, and moved Catastrophe to the top of the list. Cromwell trotted in wearing the second apron. Bill pointed at him with renewed hope. Molly slowly capped her pen. The small terrier, she decided, could be Professor Marvo.
created_at
"2026-10-27T10:40:44.267Z"
10/27/2026, 10:40:44 AM
model
"moonshotai/kimi-k3.1-pro-20261023"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1647
moderation_latency
(null)
generation_time
241666
tokens_prompt
75
tokens_completion
7787
native_tokens_prompt
146
native_tokens_completion
7993
native_tokens_completion_images
(null)
native_tokens_reasoning
7329
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
0.080222
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.1; linux; x64))"
http_referer
(null)
request_id
"req-1793097644-84wGwX5cza1OtTTPEsPL"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1793097644-sV0RyaudZ6KzO0vZ0Lek"
upstream_id
"chatcmpl-6ae07fadb7033f358251bb87"
provider_responses
0
endpoint_id
"a3fafaac-f097-4ef1-86b3-842854dad5c6"
id
"chatcmpl-6ae07fadb7033f358251bb87"
is_byok
false
latency
1647
model_permaslug
"moonshotai/kimi-k3.1-pro-20261023"
provider_name
"Moonshot AI"
status
200
total_cost
0.080222
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
71.32%
Dialogue to Total Word Ratio
Ratio: 32.60%, Deviation: 2.60%
neededClean
false
wordsTotal
500
wordsDialogue
163
85.6598%