NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5GS0VJNSSMPZT2HFQKGWW
Run Details
Model:
DeepSeek V3.2
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
13.6s
Cost:
$0.000246
Provider:
SiliconFlow
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
Molly traced a chubby crayon line across the map, which was, in fact, a brightly illustrated invitation to Timmy’s Sixth Birthday. Bill adjusted his spectacles, squinting at the rendered floorplan of the suburban living room. “The primary target cluster,” Molly whispered, tapping a cluster of ‘X’s near the fireplace, “will be here post-cake. The pinata is a tactical complication.” Bill nodded, his expression grave. He pointed a stubby pencil at a drawing of a man with a red nose. “The hired entertainment is our window. When the clown begins his tragic balloon animal phase, the adult guardians will be incapacitated by nostalgia and mild horror. That’s our ingress.” Molly leaned back, surveying their work. The plan was flawless, a masterpiece of low-stakes larceny. They had cased the joint, noting the schedule for musical chairs and the predictable sugar-crash that would follow. “We go in as caterers,” Bill stated. “No,” Molly corrected, a sly smile spreading. “As the magician’s assistants. No one questions people carrying a large, empty box.” She folded the map with reverence. All that stood between them and a haul of action figures and glittery notebooks was a battalion of six-year-olds. It was, they both agreed, their most daring caper yet.
created_at
"2026-02-16T21:26:03.770338+00:00"
2/16/2026, 9:26:03 PM
model
"deepseek/deepseek-v3.2-20251201"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1651
moderation_latency
(null)
generation_time
13516
tokens_prompt
77
tokens_completion
644
native_tokens_prompt
67
native_tokens_completion
542
native_tokens_completion_images
(null)
native_tokens_reasoning
260
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
0.00024573
router
(null)
provider_responses
0
id
"019c685865b10a90a5657545c5506466"
status
200
is_byok
false
latency
1651
endpoint_id
"b39a588a-b338-43ba-b046-226b9e24afa8"
provider_name
"SiliconFlow"
model_permaslug
"deepseek/deepseek-v3.2-20251201"
api_type
"completions"
id
"gen-1771277149-UYBI53VdLfh5COKsyDqn"
upstream_id
"019c685865b10a90a5657545c5506466"
total_cost
0.00024573
cache_discount
(null)
upstream_inference_cost
0
provider_name
"SiliconFlow"
Evaluation details
Result
Evaluator
Details
Meta Data
95.9868%
Matches word count
n/a
neededClean
false
words
208
0.0036%
Dialogue to Total Word Ratio
Ratio: 27.88%, Deviation: 17.88%
neededClean
false
wordsTotal
208
wordsDialogue
58
47.9952%