NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5N1VZZKRB5GWZS3MNGXCT
Run Details
Model:
DeepSeek V3.2
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
11.1s
Cost:
$0.000154
Provider:
Novita
Model input
User/Human
You are a author of comedic heist books. Please write a 200 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
The warehouse loft smelled of dust and desperation. Molly stared at the crayon-smeared schematic. "It's a hard target, Bill. Perimeter is soft—balloon arch, bouncy castle—but the interior is a nightmare. Thirty miniature adversaries, hopped on sugar and lemonade." Bill polished a grappling hook fashioned from a candy cane. "Our intel says the loot is centralized under a My Little Pony-themed 'Tree of Giving.' A classic snatch-and-grab." "During the piñata frenzy? That's our window. But the exfil is tight. You'll be burdened with bags of plush toys and mediocre building blocks." "I can handle the weight. What's the primary complication?" "The Asset." Molly tapped a photo of a shaggy pony-ride operator. "Clyde. He's ex-carnival. Sees everything. We need a diversion." Bill grinned, holding up a whoopee cushion and a walkie-talkie. "I've procured a multi-phase auditory distraction. Phase One: Deploy the cushion near the cake. Phase Two: When chaos peaks, I hit the 'tree.' You keep Clyde occupied." "By?" "Ask him about the existential loneliness of the carousel horse. It'll break him." Molly nodded, folding the map. "Alright. We go in at 1400 hours, during the 'Musical Chairs' meltdown. Remember, no witnesses." "Understood. And Molly? If we get cornered... we cry. It’s a kids' party. They’ll have to forgive us." "Improvisation. I like it."
created_at
"2026-02-16T21:28:21.255812+00:00"
2/16/2026, 9:28:21 PM
model
"deepseek/deepseek-v3.2-20251201"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1082
moderation_latency
(null)
generation_time
11041
tokens_prompt
77
tokens_completion
359
native_tokens_prompt
67
native_tokens_completion
341
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)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.000154423
router
(null)
provider_responses
0
id
"b41752e5d816e38430dd6be6b44d0565"
status
200
is_byok
false
latency
1082
endpoint_id
"8a5fb6d3-ef71-4958-af54-269ffa3a3c6e"
provider_name
"Novita"
model_permaslug
"deepseek/deepseek-v3.2-20251201"
api_type
"completions"
id
"gen-1771277289-uSGzcxxjwUpiFS6CokQf"
upstream_id
"b41752e5d816e38430dd6be6b44d0565"
total_cost
0.000154423
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Novita"
Evaluation details
Result
Evaluator
Details
Meta Data
20.1897%
Matches word count
n/a
neededClean
false
words
220
0.0000%
Dialogue to Total Word Ratio
Ratio: 76.82%, Deviation: 26.82%
neededClean
false
wordsTotal
220
wordsDialogue
169
10.0948%