NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KAGA2GC3VD78CJFWVHTYWS
Run Details
Model:
DeepSeek V4 Pro 0813 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 24s
Cost:
$0.016999
Provider:
Sail Research
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 leaned over the diner table. "Okay, run it again." Bill sighed. "It's a children's party, Molly. We walk in, we take the gift table, we walk out." "Too simple. Kids have eyes. Parents have phones. We need a cover." "Like what, a magician?" "Exactly. I bought a cape." "You can't do magic." "I don't need magic. I need distraction." Molly spread her hands. "I yell 'free pony rides outside,' every kid stampedes, we grab the presents." "There's no pony outside." "They'll figure that out after we're gone." "And the parents? They don't stampede." "They follow their kids. It's biology." Bill tapped the table. "What about the birthday mom? She guards the gift table like a hostage negotiator." "That's your job. You're the cake inspector." "There's no such thing." "There is today. You wear a white coat, carry a clipboard, say the frosting is a safety hazard, and ask her to show you the kitchen." "And then what?" "I steal." Bill considered. "What if a kid asks me to do a magic trick?" "Tell him you make presents disappear." "That's just confessing." "Kids love honesty." Molly pulled out a napkin with a crayon map. "The gift table is in the corner by the juice boxes. We enter at two p.m. during the piñata." Bill pointed. "Why during the piñata?" "Because every adult is praying a kid doesn't get hit with a bat. It's the only time nobody watches the presents." "Except the birthday kid. He'll watch his gifts like a tiny dragon." Molly smiled. "Bribe him with the biggest juice box." "That's not a bribe; that's hydration." "Fine. We tell him the piñata is full of candy. It's not. It's full of marbles." Bill frowned. "Why marbles?" "Chaos. Marbles on a lawn, parents slip, we exit through the side gate." "Slipping parents is a lawsuit, Molly." "We're already committing a felony; a little personal injury adds character." Molly folded the napkin. "We need a getaway vehicle." "I was thinking a minivan with a 'Baby on Board' sticker." "Perfect. Nobody suspects a minivan. We'll pull up with the sliding door open." Bill nodded slowly. "And if the cops show up?" "Then we're just two clowns escaping a birthday. Honk the horn twice." "I'm not honking." "Then we'll be arrested, but at least the kids will think it's street theater." She walked to the door. "Bring a gift bag. Empty. And a lollipop for the tiny dragon." Bill followed. "What's the lollipop for?" "Insurance." "Insurance against what, a toddler uprising?" "Worse. A toddler with a whistle." "You're afraid of a whistle?" "I'm afraid of a whistle in a balloon animal, Bill. That thing can summon parents from three blocks." Bill stopped at the car. "So we bring earplugs?" "No. We bring a second lollipop. One to bribe, one to glue his lips shut." "You can't glue a kid's lips shut with candy." "Not with that attitude." Molly unlocked the van. "Also, we need a fake birthday cake." "Why?" "For the cake inspector. You can't inspect a cake that doesn't exist." Bill groaned. "So now I'm baking?" "Don't be ridiculous. We steal one from the supermarket. And if anyone asks, it's a gift for the birthday prince." "His name is Kevin." "Then the cake says 'Happy Birthday Kevin.' I'll write it in frosting in the parking lot." "You'll get pulled over for defacing a cake." "I'll do it while you drive. Steady hands." Bill started the engine. "I hate that this is starting to make sense." "That's the sugar rush talking." Molly buckled her seatbelt. "Step on it. The piñata won't swing itself."
created_at
"2026-09-03T09:44:53.435Z"
9/3/2026, 9:44:53 AM
model
"deepseek/deepseek-v4-pro-20260813"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
618
moderation_latency
(null)
generation_time
72357
tokens_prompt
75
tokens_completion
4145
native_tokens_prompt
146
native_tokens_completion
4244
native_tokens_completion_images
(null)
native_tokens_reasoning
3377
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
0.01699896
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788428693-5qRH5ZtPvWkHG2QGr9T1"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788428693-kNaclpjFSY56CJIZfEbV"
upstream_id
"asapreq_028733cf702ab9431e4103635f4cf1f6"
provider_responses
0
endpoint_id
"ca5f0ff9-8764-4809-b5fa-968219bdfc4e"
is_byok
false
latency
10075
model_permaslug
"deepseek/deepseek-v4-pro-20260813"
provider_name
"NextBit"
status
504
1
endpoint_id
"6811c06b-8624-4796-ad6e-807e33b97ae6"
id
"asapreq_028733cf702ab9431e4103635f4cf1f6"
is_byok
false
latency
618
model_permaslug
"deepseek/deepseek-v4-pro-20260813"
provider_name
"Sail Research"
status
200
total_cost
0.01699896
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Sail Research"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
600
0.00%
Dialogue to Total Word Ratio
Ratio: 89.00%, Deviation: 19.00%
neededClean
false
wordsTotal
600
wordsDialogue
534
0.0000%