NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KCK26DHKEHMGH286R2FA69
Run Details
Model:
DeepSeek V4 Pro 0813 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 50s
Cost:
$0.046524
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
96.11%
Dialogue to Total Word Ratio
(details)
98.0533%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 hand-drawn map across the sticky diner table, smoothing its wrinkles with the edge of a ketchup bottle. The map depicted a suburban backyard, a bouncy castle, and a unicorn piƱata labelled "possible guard." Bill leaned in so close his nose almost touched the gift table marked in red crayon. "The target is a six-year-old's birthday party," Molly said. "High-value presents, minimal security." Bill nodded. "What's the play?" Molly tapped the map. "We enter as clowns." Those five words were enough to make Bill's eyes widen. He imagined the oversized shoes, the red noses, the honking. He also remembered the last time he wore a wig. It had ended with a security guard, a runaway pony, and a very angry grandmother. But the promise of unguarded presents drew him back. "We slip in during the magician's set," Molly continued. "Parents will be filming. Kids will be distracted. We load the gifts into a wagon disguised as a bubble machine." Bill frowned at the map. He traced the path from the gate to the gift table with a fry. "What about the birthday girl?" "Glitter cannon," Molly said. "Stay clear." Bill cracked his knuckles. "And the grandma?" Molly sighed. "She watches the table like a hawk. We'll need a distraction." She pulled two lollipops from her coat and laid them beside the map as if they were tactical grenades. Emergency bribes, she explained, for any child who started crying. The chocolate ones were reserved for themselves. Bill tucked one behind his ear and resumed studying the layout. The plan had a certain elegance. They would arrive as party clowns, wait for the magician's grand illusion, and then quietly transfer the presents into their customised wagon. Shiny paper indicated premium goods; lumpy wrapping meant hand-knitted nightmares from Aunt Carol. Cash envelopes from grandparents would be a bonus. Bill's expression turned serious. "We need a getaway vehicle." Molly considered this. "My nephew's wagon." "Too conspicuous." "My cousin's stroller." Bill shook his head. "Cup holders?" "All-terrain wheels." They locked eyes. The stroller was perfect. Outside the diner, a minivan rolled past with a Baby on Board sticker stuck to the rear window. Molly pointed at it with the reverence of a bank robber spotting an armour truck. The sight seemed to confirm their mission. Birthdays were big business, and they intended to collect. "One condition," Bill said. Molly raised an eyebrow. "We split the good chocolate." She squirted him with the fake flower. "Deal." Bill wiped his cheek with a paper napkin. "So much for trust." Molly rolled up the map, tucked the lollipops into her sock, and stood. "Remember the code word?" Bill tapped his red nose. "Honk twice for trouble." "And once for cake?" "Obviously." "Then tomorrow we become clowns." "Or felons with glitter in our shoes." "Same thing. Now let's practise our honking." Bill sighed. "I'm going to need a juice box." "Bring two," Molly said. "Glitter gets in your throat." "The real score is the juice box."
created_at
"2026-09-03T10:21:18.942Z"
9/3/2026, 10:21:18 AM
model
"deepseek/deepseek-v4-pro-20260813"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
960
moderation_latency
(null)
generation_time
169677
tokens_prompt
75
tokens_completion
12947
native_tokens_prompt
146
native_tokens_completion
13773
native_tokens_completion_images
(null)
native_tokens_reasoning
13080
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"
service_tier
(null)
usage
0.04652373
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788430878-6XrQc1rc5BbBHC5qkqwZ"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788430878-KXumXuLebS6LDEouKL2u"
upstream_id
"chatcmpl-ccaf2f75-5301-911d-b973-bb54882862e2"
provider_responses
0
endpoint_id
"e719d276-b080-47ee-ae4f-63ed656a0b7a"
id
"chatcmpl-ccaf2f75-5301-911d-b973-bb54882862e2"
is_byok
false
latency
960
model_permaslug
"deepseek/deepseek-v4-pro-20260813"
provider_name
"Alibaba"
status
200
total_cost
0.04652373
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Alibaba"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
499
96.11%
Dialogue to Total Word Ratio
Ratio: 30.89%, Deviation: 0.89%
neededClean
false
wordsTotal
505
wordsDialogue
156
98.0533%