NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2GFMWKY90VXYG1YMDYEFTW
Run Details
Model:
DeepSeek V4 Pro (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
27.1s
Cost:
$0.000956
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0001%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 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 the blueprints across the sticky surface of the diner table, right over a syrup ring that glued the paper to the Formica. She tapped a chubby finger on a hand-drawn clown face she’d scrawled in the margin of the Happy Time Party Palace floor plan. A dollop of scrambled egg trembled on her lower lip. “Okay, Bill, talk me through the part where the guy in the inflatable unicorn suit is supposed to be inconspicuous,” she said. Bill leaned back and cracked his knuckles with the serene confidence of a man who has never once correctly anticipated consequences. “It’s multi-phase, Moll. The unicorn is essential. According to my surveillance, the moment the hired princess arrives, all parental cognition switches off. They see a glittery horn and they go into a sort of diabetic coma of trust. I waltz in, collect the gift table haul into a reinforced wheelie bin disguised as a giant cupcake, and waltz out.” “You can’t waltz in a unicorn suit. It has hooves. You’ll shuffle.” “I’ll shuffle out. Semantics. The point is, you’re creating the diversion that makes the shuffle possible.” He stole a triangle of her toast. Molly snatched it back and bit it precisely in half. “By being a magician. A magician with a live rabbit. Bill, I’m allergic to rabbits. My eyes will swell shut. I’ll be a blind magician pulling endless scarves from my own weeping face.” “That’s pathos. Kids love pathos. A sad clown is a memorable clown.” He waved away her objection. “The rabbit is non-negotiable. I already rented him from a guy in the classifieds. His name is Houdini Muffincrumb and he comes with his own tiny top hat. You can’t disappoint an animal in evening wear.” Molly stared at him, her expression a meticulous blank. “Did you just say the rabbit’s name is Houdini Muffincrumb?” “It was a package deal. Look, while you’re palming coins and making balloon poodles go inexplicably limp, wee Madison and Aidan will be hypnotised. I’ll harness the elemental power of the unicorn to blend in with the bouncy castle landscape. The presents will vanish. We regroup in the back alley by the dumpster that smells of old sheet cake. Clean, professional, adorable.” “There’s a flaw,” Molly said, drawing a heart around the fire exit with her fork. “A six-year-old’s birthday isn’t a casino vault. The loot is forty boxes of kinetic sand and a scooter.” Bill’s eyes gleamed. “Ah, but my reconnaissance indicates the Grossman child is attending. His parents own that artisanal toy store. Their gifts will be handcrafted wooden balance bikes. Do you know what those retail for? Enough to afford a lot of antihistamines for your rabbit-induced trauma.” Molly considered this. A waitress refilled their coffee and gave the blueprints a long, incurious glance. After she left, Molly leaned forward, her voice dropping to a conspiratorial murmur. “Fine. I’ll do the magic. But I’m wearing the outfit with the cape, and I’m not honouring the rabbit’s professional dignity. If I start to swell, I’m stuffing Houdini Muffincrumb into a false-bottomed hat and directing the children to pummel the piñata until their tiny fists get tired.” Bill extended a hand across the ketchup-smudged crayon drawing of a pony on the kids’ menu. “Partner, that’s the most beautiful thing you’ve ever said. Unleash the chaos. I’ll be the mythological beast of burden.” They shook, and the unicorn suit in its black duffel bag slid off the adjacent booth and landed on the floor with a soft, hollow thud, like a dream dying in polyester.
created_at
"2026-04-25T14:25:17.721Z"
4/25/2026, 2:25:17 PM
model
"deepseek/deepseek-v4-pro-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1119
moderation_latency
(null)
generation_time
27047
tokens_prompt
77
tokens_completion
1177
native_tokens_prompt
67
native_tokens_completion
1065
native_tokens_completion_images
(null)
native_tokens_reasoning
209
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"
usage
0.000955695
router
(null)
provider_responses
0
endpoint_id
"d084170b-dd99-4e53-9ae1-b63881f3cfbf"
id
"76be85fc-ea83-48dd-b8ab-3060e7913a78"
is_byok
false
latency
270
model_permaslug
"deepseek/deepseek-v4-pro-20260423"
provider_name
"DeepSeek"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777127117-AAOMumX3i6LqFxnhoH8M"
session_id
(null)
api_type
"completions"
id
"gen-1777127117-eFUSQf7Rk4zC3nmIpO4k"
upstream_id
"76be85fc-ea83-48dd-b8ab-3060e7913a78"
total_cost
0.000955695
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
594
0.00%
Dialogue to Total Word Ratio
Ratio: 66.22%, Deviation: 16.22%
neededClean
false
wordsTotal
601
wordsDialogue
398
0.0001%