NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5TDZR52PJP2MSM438A0BT
Run Details
Model:
DeepSeek V3.2
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
32.9s
Cost:
$0.000646
Provider:
SiliconFlow
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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 spread the blueprint on the kitchen table. "Okay, Bill. The target is a six-year-old's birthday party. Prime time: during the cake chaos." Bill frowned, tapping the paper. "Stealing from kids? That's low, even for us." "Think of it as redistribution. These kids get too many toys. We're doing them a favor, teaching life lessons early." "What's the exit strategy? Through the bouncy castle?" "Exactly! We grab the presents, dive into the castle, and bounce out the back gate. The parents will be too busy singing 'Happy Birthday' to notice." "But what about the clown? Clowns are creepy. And observant." "I've handled clowns before. I'll distract him with a whoopee cushion. You load the loot into the piñata we hollowed out." "A piñata? That's genius!" "Of course. Now, remember: avoid the dog. It's a Chihuahua named Mr. Snuggles, but it has a mean bite." Bill sighed. "I can't believe I'm agreeing to this. What if we get caught?" "Then we say we're from the present recycling program. Now, let's go over the nap schedule again. The toddler goes down at two, which gives us a ten-minute window before the magician starts." "Magician? You didn't mention a magician!" "Relax. He's an amateur. His best trick is making a handkerchief disappear. We'll be fine." Bill rubbed his temples. "This is the most complicated job we've ever pulled." "And the most profitable. Think of all those unopened LEGO sets. Now, put on your party hat. We need to blend in." "What about the noise? Wrapping paper is notoriously crinkly." "We'll use the goody bags as silencers. Stuff them as we go. And no eating the candy, Bill. That's operational funds." "Fine. But if I see a pony, I'm leaving. Ponies judge."
created_at
"2026-02-16T21:31:39.313505+00:00"
2/16/2026, 9:31:39 PM
model
"deepseek/deepseek-v3.2-20251201"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2907
moderation_latency
(null)
generation_time
32655
tokens_prompt
77
tokens_completion
1666
native_tokens_prompt
67
native_tokens_completion
1495
native_tokens_completion_images
(null)
native_tokens_reasoning
1071
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
0.00064599
router
(null)
provider_responses
0
id
"019c685d38df60f810cf0f596c0e830e"
status
200
is_byok
false
latency
2907
endpoint_id
"b39a588a-b338-43ba-b046-226b9e24afa8"
provider_name
"SiliconFlow"
model_permaslug
"deepseek/deepseek-v3.2-20251201"
api_type
"completions"
id
"gen-1771277465-4Fzp83GmCl4kNHMG55w9"
upstream_id
"019c685d38df60f810cf0f596c0e830e"
total_cost
0.00064599
cache_discount
(null)
upstream_inference_cost
0
provider_name
"SiliconFlow"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0000%
Matches word count
n/a
neededClean
false
words
291
98.5593%
Dialogue to Total Word Ratio
Ratio: 93.47%, Deviation: 3.47%
neededClean
false
wordsTotal
291
wordsDialogue
272
49.2797%