NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM61BSE6GM33PZZNZHJHW8F
Run Details
Model:
DeepSeek V3.2
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
2m 32s
Cost:
$0.003523
Provider:
SiliconFlow
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
Molly smoothed the crayon-drawn map across the splintered picnic table. Bill eyed it skeptically, munching a stolen cookie. The backyard was quiet, save for distant lawnmower sounds. “The target,” Molly announced, “is little Timmy’s birthday. Guards include a dad burning burgers and a clown on a sugar high.” Bill sprayed crumbs. “A clown? I’m out.” “Don’t be a baby. We strike during musical chairs. You’ll cause a diversion by claiming you saw a celebrity.” “What celebrity?” “Someone believable. Like the local weatherman. I’ll grab the presents while they’re distracted.” “How?” “I’ll wear a piñata costume. They’ll think I’m part of the entertainment.” Bill frowned. “Stealing from kids? That’s cold.” “It’s a victimless crime! They’ll get more presents. We’re teaching them about loss early.” “And the loot?” “Mostly plastic junk, but there’s always cash from grandparents. We’ll fence the good toys online.” “What about the clown?” “We’ll exploit his fear of water balloons. I’ve got a squirt gun.” “This is the dumbest plan we’ve ever had.” “But it’ll work. Now, put on your happy face. We’re going to a party.” Bill forced a smile, which looked more like a grimace. He adjusted his shirt, already feeling ridiculous. Outside, children’s laughter floated over the fence, unknowingly inviting the least threatening heist in history.
created_at
"2026-02-16T21:37:25.598481+00:00"
2/16/2026, 9:37:25 PM
model
"deepseek/deepseek-v3.2-20251201"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1004
moderation_latency
(null)
generation_time
151475
tokens_prompt
77
tokens_completion
8815
native_tokens_prompt
67
native_tokens_completion
8344
native_tokens_completion_images
(null)
native_tokens_reasoning
8005
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
0.00352257
router
(null)
provider_responses
0
id
"019c6860b11df8faff14efc84e7d846d"
status
200
is_byok
false
latency
1004
endpoint_id
"b39a588a-b338-43ba-b046-226b9e24afa8"
provider_name
"SiliconFlow"
model_permaslug
"deepseek/deepseek-v3.2-20251201"
api_type
"completions"
id
"gen-1771277692-gVUZbDmfFCPh0Rl6KTs7"
upstream_id
"019c6860b11df8faff14efc84e7d846d"
total_cost
0.00352257
cache_discount
(null)
upstream_inference_cost
0
provider_name
"SiliconFlow"
Evaluation details
Result
Evaluator
Details
Meta Data
81.2727%
Matches word count
n/a
neededClean
false
words
212
0.0011%
Dialogue to Total Word Ratio
Ratio: 68.40%, Deviation: 18.40%
neededClean
false
wordsTotal
212
wordsDialogue
145
40.6369%