NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2EMSZCZNEN9ZV70614HKAS
Run Details
Model:
DeepSeek V4 Pro
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
17.7s
Cost:
$0.000319
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
0.01%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0049%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 venue was, as Molly had noted with professional disgust, a soft-play centre named Kingdom of Fun. Bill studied the floor plan, which was less an architectural blueprint and more a biohazard map of moated ball pits and padded slime-green volcanoes. He tapped a sausage-sized finger on the final chamber. “The loot’s in the Throne Room. Past the Toddler Gulag.” Molly, sharpening a bespoke titanium cake knife, paused to polish off a stolen canapé shaped like a pirate’s eyeball. “The silent alarm?” she asked, licking her thumb. “A sleeping mother with a monitor. One false squeak from a rubber duck and she’ll be on us with the feral speed of a parent who hasn’t slept since 2019.” Molly nodded grimly. The job required them to be invisible, blending with the chaos until the ritual sacrifice of the piñata. Their cover: children’s entertainers. Bill, a self-proclaimed cynic, would be Sparkles the Existential Clown. Molly, a woman who once hotwired a tank, would pose as a fairy princess called Glimmer, whose dodgy wing-harness concealed a foldable duffel bag of industrial capacity. “The birthday boy turns four precisely at three o’clock,” she whispered. “That’s when the sugar crash hits the horde. Paralysis by fondant. We grab the brightly coloured plastic gold and vanish into the car park fog of weeping toddlers.” Bill cracked a single, non-biodegradable balloon from his clown pouch in solemn agreement.
created_at
"2026-04-25T13:53:16.171Z"
4/25/2026, 1:53:16 PM
model
"deepseek/deepseek-v4-pro-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1013
moderation_latency
(null)
generation_time
11078
tokens_prompt
77
tokens_completion
375
native_tokens_prompt
67
native_tokens_completion
333
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
0.000318855
router
(null)
provider_responses
0
endpoint_id
"d084170b-dd99-4e53-9ae1-b63881f3cfbf"
id
"535efb27-4b88-4115-be39-2452855b2bf4"
is_byok
false
latency
279
model_permaslug
"deepseek/deepseek-v4-pro-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-1777125196-CSvq3IYE9E3flNKqGuj0"
session_id
(null)
api_type
"completions"
id
"gen-1777125196-3EFL8S4zOL36QufUPnhb"
upstream_id
"535efb27-4b88-4115-be39-2452855b2bf4"
total_cost
0.000318855
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.01%
Matches word count
n/a
neededClean
false
words
231
0.00%
Dialogue to Total Word Ratio
Ratio: 33.76%, Deviation: 23.76%
neededClean
false
wordsTotal
237
wordsDialogue
80
0.0049%