NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2FYHT8H1WHK8GARG283CRH
Run Details
Model:
DeepSeek V4 Flash
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
10.9s
Cost:
$0.000208
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
27.17%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
13.5829%
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 balanced on the edge of a wobbly garden chair, peering through binoculars across the hedge. “Okay, Bill, talk to me. What are we looking at?” Bill, crouched behind a plastic flamingo, adjusted his earpiece. “Target’s at twelve o’clock. The bounce house. Orange and green stripes. Looks like a giant, sweaty citrus fruit.” “I can see the bounce house, genius. I meant the presents. Are they guarded?” “Guarded? By a pack of seven-year-olds hopped on sugar and a woman in a cat-ears headband who keeps yelling ‘Nathan, stop licking the piñata.’ So yeah, heavily guarded.” Molly lowered the binoculars. “We need a distraction. Something loud. Something that will make every parent yank their kid toward the exits.” “I’ve got the rocket launcher in the trunk.” “We are not using the rocket launcher, Bill. That thing’s for security systems, not toddler parties.” “Fine. How about a fake bee attack? I bought a drone with a buzzing attachment.” “That’s stupid. Children love bees. They’ll try to catch it.” Bill sighed, shifting his weight. “What’s your brilliant plan, then?” “We steal the cake first.” “The cake?” He stood up, immediately ducking as a stray frisbee sailed over his head. “We’re here for the presents. The wrapped ones with the shiny bows and the gift cards from Grandma.” “Yes, but if the cake vanishes, everyone panics. Parents start blaming each other. ‘Who ate the Elsa cake, Brenda?’ Chaos, Bill. Pure chaos. And in the chaos, I slip the loot bag under my jacket.” Bill considered this, rubbing his chin. “What if they have a backup cake?” “In this economy? No one has a backup cake.” She climbed off the chair, brushing grass from her knees. “You saw the spreadsheet. One cake. Vanilla with buttercream. Serves twenty-five. We steal it, the party dissolves, and while they’re all crying into their fruit punch, we grab the pile of presents.” “And where do we stash the cake?” “In the car. We can eat it later. It’s a bonus.” Bill grinned, a slow, lopsided thing. “You know, Moll, sometimes your mind is a beautiful, terrifying machine.” “Flattery gets you nowhere. Now, here’s the play. You approach from the back, near the grill. The dad with the tongs is distracted by the hot dogs. You grab the cake, run through the bounce house—it’ll buy you speed—and head for the fence.” “Through the bounce house? I’ll look ridiculous.” “You’ll look like a criminal in a hurry. Get ready. I’ll mingle. Fake a phone call. ‘Oh no, the sprinklers are on at my house!’ Standard stuff.” Bill nodded, then hesitated. “What if the birthday kid cries?” “Then we leave a five-dollar bill in the gift bag. It’s a write-off.” He laughed, low and quiet. “You’re a monster.” “I’m a professional, Bill. There’s a difference.” She adjusted her sunglasses. “On my count. Three… two… go make children weep for their sugar.”
created_at
"2026-04-25T14:15:57.516Z"
4/25/2026, 2:15:57 PM
model
"deepseek/deepseek-v4-flash-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
815
moderation_latency
(null)
generation_time
10665
tokens_prompt
77
tokens_completion
812
native_tokens_prompt
67
native_tokens_completion
711
native_tokens_completion_images
(null)
native_tokens_reasoning
0
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
0.00020846
router
(null)
provider_responses
0
endpoint_id
"722e0746-2034-4d6e-a08f-e6271849ee9e"
id
"9b08ed80-ec81-429e-949f-ceb128b5869c"
is_byok
false
latency
267
model_permaslug
"deepseek/deepseek-v4-flash-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-1777126557-7KuQx9AzbUo7mHnKVXfY"
session_id
(null)
api_type
"completions"
id
"gen-1777126557-peypIh1LORjHM7pMCfsb"
upstream_id
"9b08ed80-ec81-429e-949f-ceb128b5869c"
total_cost
0.00020846
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
27.17%
Matches word count
n/a
neededClean
false
words
481
0.00%
Dialogue to Total Word Ratio
Ratio: 83.16%, Deviation: 33.16%
neededClean
false
wordsTotal
487
wordsDialogue
405
13.5829%