NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5XYNWHSH7P0RC2BNX8NHF
Run Details
Model:
DeepSeek V3 (2025-03-24)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
28.4s
Cost:
$0.000605
Provider:
ModelRun
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 adjusted the fake mustache and frowned at Bill. "You sure this is the best disguise? Feels a bit... obvious." Bill adjusted the matching mustache and grinned. "Nobody suspects a clown with a mustache, Mol. It’s psychological warfare. They’ll be too busy wondering why we have mustaches to notice we’re stealing." Molly sighed. "Or they’ll wonder why two clowns look like they’re smuggling caterpillars on their faces." She peered through the binoculars at the backyard party below. Balloons, cake, and a piñata shaped like a unicorn—guarded by a horde of sugared-up six-year-olds. "I still think the hard part’s gonna be getting past Tiny Tim over there." Bill followed her gaze to a particularly ferocious toddler pile-driving another into the grass. "Eh. Kid chaos works in our favor. Distraction central." He tapped the blueprint spread between them. "Now, the presents are all stacked by the bouncy castle. Simple grab-and-go." "Simple? There’s, like, twenty hyperactive sentinels between us and the loot." "That’s why we’ve got *this*." Bill pulled a jumbo bag of candy from his duffel. Molly groaned. "You’re bribing them? That’s your masterstroke?" "It’s not a bribe. It’s *tactical redistribution.*" He shook the bag. "I toss theวรรณ candy east, kids stampede east. You grab the goods west. Boom." She stared. "That only works if none of them have object permanence." "That’s why they’re perfect targets. They think peek-a-boo is dark magic." A scream pierced the air as Tiny Tim mounted another child like a rodeo. Molly rubbed her temples. "Okay, fine. But if this goes south, I’m blaming you *and* your ridiculous mustache." 은 Bill saluted. "Deal. Ready for Operation: Sugar Overload?" Molley adjusted her clown shoes—twice the size of her actual feet. "Let’s just hope nobody asks us to actually *clown*." Bill scoffed. "Please. How hard can balloon animals be?" He blew up a balloon, twisted it, and—*pop*. A deflated worm flopped into his palm. Molley deadpanned. "We’re doomed." Bill chucked the candy bag into the fray. "Nah. We’re professionals." The candy rained down like manna from heaven. Tiny Tim’s head snapped up. The horde froze, then *charged*. Molley bolted for the presents. Bill tripped over his shoes and face-planted into the grass. A tiny voice piped up beside him. "Mister Clown? You okay?" Bill lifted his head, grass sticking to his mustache. "Just… clowning around, kid." Molley stuffed armfuls of gifts into her sack, then froze as a mom narrowed her eyes. "Can I help you?" Molley forced a grin. "Party… cleanup crew?" The mom crossed her arms. "The party’s not over." Bill popped up beside Molley, mustache askew. "Early birds get the… uh… presents?" The mom sighed. "Nice try." Molley dropped the sack. "Run for it!" Bill yelped, clown shoes flapping, as they bolted for the fence—giftless, but hilariously alive.
created_at
"2026-02-16T21:33:29.833603+00:00"
2/16/2026, 9:33:29 PM
model
"deepseek/deepseek-chat-v3-0324"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2616
moderation_latency
(null)
generation_time
28355
tokens_prompt
77
tokens_completion
800
native_tokens_prompt
66
native_tokens_completion
768
native_tokens_completion_images
(null)
native_tokens_reasoning
0
native_tokens_cached
65
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.00060456
router
(null)
provider_responses
0
id
"chatcmpl-879443319b5f4a79a395feccf86712c6"
status
200
is_byok
false
latency
103
endpoint_id
"a3938710-1b99-4a1f-85f4-2494f3f659d3"
provider_name
"ModelRun"
model_permaslug
"deepseek/deepseek-chat-v3-0324"
api_type
"completions"
id
"gen-1771277581-IW65eRQmuH1LBqldq0zt"
upstream_id
"chatcmpl-879443319b5f4a79a395feccf86712c6"
total_cost
0.00060456
cache_discount
(null)
upstream_inference_cost
0
provider_name
"ModelRun"
Evaluation details
Result
Evaluator
Details
Meta Data
1.0360%
Matches word count
n/a
neededClean
false
words
474
0.0000%
Dialogue to Total Word Ratio
Ratio: 47.05%, Deviation: 22.95%
neededClean
false
wordsTotal
474
wordsDialogue
223
0.5180%