NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2GX9AGT4MC5MZMQ53F4CN8
Run Details
Model:
DeepSeek V4 Flash
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
11.0s
Cost:
$0.000217
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
99.38%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.6885%
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
The rusted van smelled of old cheese and desperation. Molly, a woman whose sharp cheekbones could cut glass, was wedged between a bundle of rainbow-colored balloons and a cardboard cake with a spring-loaded snake inside. Across from her, Bill—a man built like a nervous bread loaf—was fumbling with a tablet displaying the schematics of the Wigglytown Rec Center. “Okay, run it past me again,” Molly said, tapping a fingernail against the window. “But slower, like I’m a toddler. Which, given the venue, feels appropriate.” Bill squinted at the screen. “The party is for a kid named Bradley. Eight years old. Theme is ‘Super Spies vs. The Clumsy Clowns.’ The presents are stacked in a designated ‘Grotto of Gratitude’—that’s a cardboard castle with a drawbridge.” “A drawbridge. For presents.” “Hand-painted. Rope-operated.” Molly pressed her lips together. “And our in?” “We’re the balloon artists. Ida, my sister, was supposed to cover it, but she’s got shingles. I told them I was her assistant, Milton. You’re my trainee, Geraldine.” “Geraldine.” “It was either that or ‘Broomstick Brenda.’ I thought Geraldine had more dignity.” Molly rubbed her temples. “So we walk in, twist some balloons into... what?” Bill flipped through a notebook. “Poodles. Swords. A giraffe. That’s all Ida taught me, and the giraffe looks more like a deflated camel. But they’re eight. They won’t know.” “They’ll know when I hand them a limp tube of rubber and tell them it’s a wiener dog.” “That’s the spirit.” Molly leaned forward, the balloons crinkling. “And the extraction? The Grotto of Gratitude is in the center of the room. There will be parents. There will be sticky fingers. There will be a child named Bradley who will cry if he sees a man in a fanny pack walking off with his loot.” Bill smiled, a thin, unsettling gesture. “You’re forgetting the distraction.” He pointed to a large duffel bag at his feet. Molly unzipped it. Inside was a single, oversized clown shoe, a rubber chicken, and a can of whipped cream. “That’s your plan?” she whispered. “You’re going to juggle and squirt dairy on someone?” “No,” Bill said, holding up a small remote. “I’ve wired the cake. When I hit this button, the snake pops out. But I’ve also rigged the frosting with a packet of itching powder. The kids will scratch. The parents will panic. We slip behind the castle, fill our bags, and exit through the emergency kitchen door while the birthday boy is tearing his own shirt off.” Molly stared at him. A long, quiet beat. Then she cracked a grin. “Geraldine,” she said. “I like it. But I’m doing the poodle. Your giraffe looks like a broken spine.” Bill nodded, starting the van. “Deal. And if it goes sideways?” “We blame the shingles.” The van rolled toward the rec center, where a giant inflatable bouncy castle grinned at them like a bright yellow tooth. Somewhere inside, Bradleys presents waited. And somewhere in the back of the van, a rubber chicken seemed to be silently laughing.
created_at
"2026-04-25T14:32:44.628Z"
4/25/2026, 2:32:44 PM
model
"deepseek/deepseek-v4-flash-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
611
moderation_latency
(null)
generation_time
10137
tokens_prompt
77
tokens_completion
836
native_tokens_prompt
67
native_tokens_completion
743
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
0.00021742
router
(null)
provider_responses
0
endpoint_id
"722e0746-2034-4d6e-a08f-e6271849ee9e"
id
"e8e31e68-51a9-456b-8014-1acd36e48250"
is_byok
false
latency
279
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-1777127564-KFlgPz0CLsdkFfxSzUur"
session_id
(null)
api_type
"completions"
id
"gen-1777127564-33GTmRB6Vs18TmeU7lH6"
upstream_id
"e8e31e68-51a9-456b-8014-1acd36e48250"
total_cost
0.00021742
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
99.38%
Matches word count
n/a
neededClean
false
words
505
0.00%
Dialogue to Total Word Ratio
Ratio: 60.43%, Deviation: 30.43%
neededClean
false
wordsTotal
508
wordsDialogue
307
49.6885%