NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5SKG3RM5KB67T9G6W92ZG
Run Details
Model:
DeepSeek V3.2
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
9.1s
Cost:
$0.000149
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 squinted at the blueprints, her finger tracing a line. "It's a fortified position, Bill. Perimeter is soft—balloon arches and a bouncy castle—but the target is deep inside. The 'Treasure Room.' Also known as Karen's sunroom." Bill adjusted his beanie, peering over her shoulder. "Guard details?" "Two rotating sentries. 'Uncle Derek,' prone to napping in front of the football, and 'Aunt Susan,' a hawk with a prosecco lens. The main obstacle is a miniature, hyper-vigilant security system named 'Lily,' age seven. She never leaves the treasure." "So we go in disguised. Operation Pin the Tail on the Donkey." Bill chewed his lip. "Extraction route?" "Molly, age five, cousin's friend. You're her awkward dad, 'Steve.' We blend, we mingle, we load the gift sacks into the oversized diaper bag." She tapped a photo of a rainbow-colored piñata. "Diversion. You trigger a candy avalanche at 3:15. I'll make the grab during the sugar-fueled chaos." Bill nodded, a glint in his eye. "Clean. But what's our cover if we're made?" Molly handed him a party hat. "We're devastated parents. We just really, really wanted our kid to have a good birthday." She smirked. "Now, let's rehearse your small talk about preschool soccer. It’s key to the alibi."
created_at
"2026-02-16T21:30:49.352334+00:00"
2/16/2026, 9:30:49 PM
model
"deepseek/deepseek-v3.2-20251201"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1113
moderation_latency
(null)
generation_time
9030
tokens_prompt
77
tokens_completion
339
native_tokens_prompt
67
native_tokens_completion
327
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
0.000148823
router
(null)
provider_responses
0
id
"7ab4e94c7d47f9025a2681d61f075084"
status
200
is_byok
false
latency
1113
endpoint_id
"8a5fb6d3-ef71-4958-af54-269ffa3a3c6e"
provider_name
"Novita"
model_permaslug
"deepseek/deepseek-v3.2-20251201"
api_type
"completions"
id
"gen-1771277438-t0uYO6ZEDoRUEA9VPVKR"
upstream_id
"7ab4e94c7d47f9025a2681d61f075084"
total_cost
0.000148823
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Novita"
Evaluation details
Result
Evaluator
Details
Meta Data
43.3783%
Matches word count
n/a
neededClean
false
words
217
0.0000%
Dialogue to Total Word Ratio
Ratio: 75.12%, Deviation: 25.12%
neededClean
false
wordsTotal
217
wordsDialogue
163
21.6892%