NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KA4AHJFS4SNQ683Y2Y9N7S
Run Details
Model:
DeepSeek V4 Pro 0813 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
2m 54s
Cost:
$0.041535
Provider:
GMICloud
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
99.98%
Dialogue to Total Word Ratio
(details)
99.8632%
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 crayon-drawn map across the van's sticky floor, next to a half-eaten juice box and a deflated balloon animal. The map had been drawn by a seven-year-old informant named Kevin, who accepted payment in gummy worms and never asked questions. Bill peered through a pair of plastic binoculars shaped like a giraffe. The van smelled of old socks, stale popcorn, and the faintly chemical sweetness of fruit punch powder. Outside, the community center parking lot was empty except for a minivan covered in cartoon duck stickers. Somewhere inside that building, a birthday party was reaching the sugar event horizon, and the presents were piled like pirate treasure. "We are not going in through the bouncy castle again. I still have cake in my left ear," Molly said. Bill shrugged and lowered the binoculars. "That was a tactical retreat." She tapped the map. "The presents are stacked by the pinata. A clown named Bubbles is guarding them." "He has a whistle and zero patience," Bill said. Molly said, "We go in as party entertainers. I'll be a magician." Bill winced and scratched the back of his neck. "Your last trick made a hamster appear in the birthday cake." Molly pulled a roll of stickers from her pocket. "We use these security stickers and walk through the front door." "That's insane," Bill said. "Insane is the only way through a room of toddlers on fruit punch. They count everything," Molly said. Bill considered this. "What about the clown?" Bill had once watched Bubbles fold a balloon into a poodle while maintaining unwavering eye contact with a man who tried to cut the cake early. The clown did not blink. He simply twisted a latex leg and squeezed it until it shrieked. "Spill glitter in his shoes. He'll be crying in the bathroom," Molly said. Molly folded the map into a tiny paper airplane. The crayon lines smudged on her fingers. She tossed the airplane at Bill's head, but it wobbled and landed in an open bag of cheese puffs. "The birthday boy is turning five. We grab the big purple box with the robot and three gift bags." The map's red crayon arrow pointed to a side door near the craft table. At their last job, they had escaped through a ball pit and lost two bags of loot to a collapsing slide. This time, they would be methodical. Bill squinted. "What about the sign that says no touching?" "Then we don't touch the table. We lift with our knees," Molly said. Bill sighed and picked up the giraffe suit from behind a pizza box. The suit's plastic eyes stared at the ceiling with the dead expression of a part-time mascot. Bill held the suit at arm's length. A faint cloud of dust rose from the fur. He looked at Molly, who was already smearing a security sticker onto her collar like it meant something. Somewhere, a child began to scream with joy. "Fine. Let's go."
created_at
"2026-09-03T09:38:18.817Z"
9/3/2026, 9:38:18 AM
model
"deepseek/deepseek-v4-pro-20260813"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2855
moderation_latency
(null)
generation_time
173881
tokens_prompt
75
tokens_completion
11653
native_tokens_prompt
146
native_tokens_completion
12291
native_tokens_completion_images
(null)
native_tokens_reasoning
11655
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
0.041535318
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788428298-tUPRjJoHbBJlW2KmMrMh"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788428298-fphQdlvsRmujQdfbMuBj"
upstream_id
"0217884283005343335ce5dc727daeef6b2fdeb6cd96fc297490e"
provider_responses
0
endpoint_id
"a11bdcb9-acb5-49df-9ff5-8c1486186664"
id
"0217884283005343335ce5dc727daeef6b2fdeb6cd96fc297490e"
is_byok
false
latency
2855
model_permaslug
"deepseek/deepseek-v4-pro-20260813"
provider_name
"GMICloud"
status
200
total_cost
0.041535318
cache_discount
(null)
upstream_inference_cost
0
provider_name
"GMICloud"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
99.74%
Matches word count
n/a
neededClean
false
words
496
99.98%
Dialogue to Total Word Ratio
Ratio: 29.94%, Deviation: 0.06%
neededClean
false
wordsTotal
501
wordsDialogue
150
99.8632%