NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD7WXNK7JWH9WGRER116V7Q
Run Details
Model:
MoonshotAI: Kimi K3 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
41.2s
Cost:
$0.015434
Provider:
Moonshot AI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
5.93%
Dialogue to Total Word Ratio
(details)
2.9667%
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 spread the invitation across the steering wheel like it was a treasure map drawn by a drunk giraffe. “Operation Pinata begins at three,” she said. “Crowns, cake, thirty unsupervised gift bags. In and out before the clown finishes making balloon swords.” Bill squinted at the glitter lettering. “Why does the invitation say ‘no grown-ups unless responsible’? That feels personally aimed at me.” “It is aimed at you. Last job you waved at a security camera and asked if it validated parking.” “I was being friendly. Cameras remember kindness.” “Cameras remember your face, Bill. That’s the whole camera business model.” “Fine. Disguises?” Molly held up two paper plates with holes cut in them. “We go as parents.” Bill stared. “Molly, those are plate masks.” “Exactly. No one looks twice at a parent holding a plate. Parents are invisible. They carry wipes and emotional damage.” “What’s my cover story?” “You’re Uncle Dennis. You know the birthday boy from soccer.” “What’s his name?” Molly checked. “Princess Sparkle-Rose.” “Terrific. I’ll say I taught her to slide tackle.” “Too sporty. Say you taught her to share.” Bill gasped. “At a children’s party? That’s science fiction.” Molly tapped the dashboard clock. “Focus. Gifts go in pillowcases. Not the good pillowcases, the ugly ones from the motel.” “Why ugly?” “Guilt transfers better to ugly fabric.” “I don’t think guilt works like laundry.” “You’ve never met my aunt Linda.” Bill opened the glove box and pulled out a kazoo. “Signal device.” “Absolutely not.” “If I see a magician, I kazoo once. If I see a bouncy castle, twice. If I see your ex dressed as Elsa, I kazoo the national anthem and we die nobly.” “There will be no dying. There will be brisk walking. Steal like a librarian: quiet, judgmental, excellent posture.” “What about the presents shaped like obvious puppies?” “No live puppies.” “Define live.” “Breathing, Bill.” “You always narrow the market.” Molly leaned closer. “Listen. The rule is simple: grab wrapped boxes, ignore sentimental handmade junk, never open anything that makes noise.” Bill nodded solemnly. “So we avoid tambourines, ukuleles, and children.” “Children are not presents.” “Tell that to the ones glued to the dessert table.” “They’re icing-drunk. Different species.” A beat passed. Bill lowered his plate mask. “Molly, are we bad people?” Molly considered. “We’re festive-adjacent opportunists.” “That sounds like a candle scent.” “Bad Decisions by Yankee.” Bill started the car, then stopped. “One more thing. If the birthday girl cries?” Molly’s mouth twitched. “We leave the bath bombs.” “All of them?” “Fine, half. We’re thieves, not monsters.” “And the clown?” Molly smiled like a shark in church. “The clown can keep his swords.” Bill put the kazoo to his lips. “Then let’s go ruin nap time.” “Softly,” Molly said, already reaching for the pillowcases. “Like librarians.”
created_at
"2026-07-25T18:16:47.038Z"
7/25/2026, 6:16:47 PM
model
"moonshotai/kimi-k3-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2312
moderation_latency
(null)
generation_time
41111
tokens_prompt
75
tokens_completion
1176
native_tokens_prompt
147
native_tokens_completion
1026
native_tokens_completion_images
(null)
native_tokens_reasoning
277
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
0.0154341
router
(null)
provider_responses
0
endpoint_id
"650c9b12-7b6d-444d-8d92-df8daef84d7a"
id
"chatcmpl-6a64fd7ecc9b99f0a3800653"
is_byok
false
latency
2312
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-1785003407-ambMzd4uua4V0ZmJ2qBK"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785003407-ONODKgsW9ehxiIkNA0Gh"
upstream_id
"chatcmpl-6a64fd7ecc9b99f0a3800653"
total_cost
0.0154341
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
463
5.93%
Dialogue to Total Word Ratio
Ratio: 77.52%, Deviation: 7.52%
neededClean
false
wordsTotal
467
wordsDialogue
362
2.9667%