NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD6B0EFPKHYMCV4BP4FK378
Run Details
Model:
MoonshotAI: Kimi K3 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
1m 13s
Cost:
$0.032924
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
75.16%
Matches word count
(details)
60.62%
Dialogue to Total Word Ratio
(details)
67.8881%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 50% 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 unfolded a crayon blueprint across the dashboard while Bill parked beside a bouncy castle shaped like a patriotic boot. Beyond the fence, thirty presents glittered under a banner celebrating Luca, age six, emperor of juice. “Catering uniforms,” Molly said. “We swap gifts for potatoes.” “Potatoes?” Bill whispered. “At a child’s party?” “Organic. Ethically disappointing.” A clown honked past wearing suspenders and menace. Bill clutched the wheel. “I feel bad. Kids remember faces, balloons, trauma.” “Relax. We leave one present.” “Which?” “The drum set. Justice demands witnesses suffer.” Molly checked her mustache in the mirror. “You create diversion: tragic balloon giraffe, emotional speech about gluten. I crawl under the gift table.” “On grass? That’s crawling, not heisting.” “Crawl like burglary is beneath you.” Bill produced two juice boxes of gin. “If caught, we were invited.” “By whom?” “Luca. He looks connected.” The birthday boy marched out in a paper crown, pointing at their van like a tiny detective with cake evidence. Molly ducked. Bill waved back, committing the felony of friendliness. “New plan,” she hissed. “Grab bags, blame clown.” “Which clown?” “All clowns are one clown legally.” They leapt out, tripped over the castle, apologized to a magician, and escaped with three yo-yos, a plush subpoena, and deep respect for Luca’s security stare.
created_at
"2026-07-25T17:49:31.478Z"
7/25/2026, 5:49:31 PM
model
"moonshotai/kimi-k3-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2616
moderation_latency
(null)
generation_time
73320
tokens_prompt
75
tokens_completion
2107
native_tokens_prompt
147
native_tokens_completion
2192
native_tokens_completion_images
(null)
native_tokens_reasoning
1842
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
0.0329241
router
(null)
provider_responses
0
endpoint_id
"650c9b12-7b6d-444d-8d92-df8daef84d7a"
id
"chatcmpl-6a64f71921f78ad95c5e70cd"
is_byok
false
latency
2616
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-1785001771-9LEIGv5g8fCHX0VbxO3G"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785001771-DjkuDXGVAq2Us3XhI2IY"
upstream_id
"chatcmpl-6a64f71921f78ad95c5e70cd"
total_cost
0.0329241
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
75.16%
Matches word count
n/a
neededClean
false
words
213
60.62%
Dialogue to Total Word Ratio
Ratio: 41.59%, Deviation: 8.41%
neededClean
false
wordsTotal
214
wordsDialogue
89
67.8881%