NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD7B0FFKBE3T8M6X8M97THG
Run Details
Model:
MoonshotAI: Kimi K3 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
26.1s
Cost:
$0.009794
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
17.01%
Dialogue to Total Word Ratio
(details)
8.5035%
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 crouched behind the privet hedge, binoculars trained on the McAllisters' back garden, where a bouncy castle swayed like a drunk giraffe. "Twenty-two presents, Bill. I counted. All wrapped, all unguarded, all sitting on a picnic table next to a clown." Bill peered through the hedge. "The clown's a problem." "The clown is a man in a wig named Gary who smells like crisps." "Gary's unpredictable, Molly. You don't know Gary's life." "Fine. New plan. We go in disguised." "As what?" Molly unzipped the duffel bag and produced two tutus, one tiara, and a fairy wand with a squeaky star. "No," Bill said. "You're a fairy godfather. It's thematic." "Molly, I'm six foot two with a moustache." "Fairy godfathers come in all shapes, Bill. It's a progressive neighbourhood." Bill held the tutu at arm's length like it might bite him. "And what do I do, sprinkle children?" "You distract. I grab. Standard snatch and dash, except the vault is a Peppa Pig gift bag." "What if someone asks which kid we're with?" "You're with Timothy." "Who's Timothy?" "There's always a Timothy. Nobody ever knows who Timothy belongs to. Timothy is a law of nature." Bill considered this. "That's actually sound." "Thank you. Now, the presents. Prioritize anything that rattles." "Rattles means cheap. Rattles means Lego knockoffs." "Rattles means easy to carry, Bill. We're not art thieves. We're present thieves. Manage your expectations." "What's the exit?" "Over the fence, through the Hendersons' garden, into the van." "The Hendersons have a dog." "The Hendersons have a pug. A pug, Bill. It will tire itself out barking and fall asleep mid-chase." "And if the clown clocks us?" "Then you do the thing." "I am not doing the thing." "You do the thing or we walk away with nothing but balloon animals." Bill sighed deeply, the sigh of a man who had done the thing before and would clearly do it again. "Fine. But if I juggle, you owe me." "I owe you nothing. You love juggling. You've juggled at three separate funerals." "Two. One was a wake." Molly shoved the tiara onto his head. "Right. We move in ten. Remember, you're excited about the party. You've been looking forward to this party all week." "I have?" "You're a fairy godfather who loves parties and definitely has a Timothy." Bill stood, tutu rustling ominously, and squared his shoulders. "Let's go rob a seven-year-old." "That's the spirit," Molly said, and together they marched toward the bouncy castle, heads high, dignity elsewhere.
created_at
"2026-07-25T18:07:00.09Z"
7/25/2026, 6:07:00 PM
model
"moonshotai/kimi-k3-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2212
moderation_latency
(null)
generation_time
25644
tokens_prompt
75
tokens_completion
663
native_tokens_prompt
147
native_tokens_completion
650
native_tokens_completion_images
(null)
native_tokens_reasoning
18
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
0.0097941
router
(null)
provider_responses
0
endpoint_id
"650c9b12-7b6d-444d-8d92-df8daef84d7a"
id
"chatcmpl-6a64fb338611480fa8ef4954"
is_byok
false
latency
2212
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-1785002820-NCPuqAs4XwD5Bf38lGUV"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785002820-QFwJGPithr34xbV0aY95"
upstream_id
"chatcmpl-6a64fb338611480fa8ef4954"
total_cost
0.0097941
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
0.00%
Matches word count
n/a
neededClean
false
words
416
17.01%
Dialogue to Total Word Ratio
Ratio: 75.95%, Deviation: 5.95%
neededClean
false
wordsTotal
420
wordsDialogue
319
8.5035%