NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5GSSRBCFBNN79VXEPMTAP
Run Details
Model:
DeepSeek V3 (2024-12-26)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
10.8s
Cost:
$0.000319
Provider:
DeepInfra
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 leaned over the table, her eyes gleaming. "Alright, Bill, here’s the plan. We hit the Johnson’s kid’s party tomorrow. They’re rich, and their brat’s turning six. Prime loot." Bill scratched his chin. "A kid’s party? What’re we stealing, crayons and juice boxes?" "Presents, Bill. Expensive ones. Think LEGO sets, iPads, and whatever trendy plastic junk parents waste money on." Molly grinned. "And the best part? No security. Just parents distracted by sugar-high brats." Bill frowned. "What if someone catches us?" "We’re dressed as clowns. Nobody questions clowns. Besides, if things go south, we toss glitter and run. Works every time." Bill snorted. "Glitter? That’s your emergency plan?" "Yes, glitter. It’s chaos in a bottle. You throw it, and everyone panics. Ever tried cleaning glitter out of a carpet? It’s a nightmare." Bill shrugged. "Alright, clown suits and glitter. But what about the kid? Won’t they notice their presents disappearing?" "Kid’s six. We distract them with a balloon animal and a lollipop. Trust me, they’ll forget their own name." Bill chuckled. "You’re diabolical. Fine, let’s do it. But if I end up covered in frosting, I’m blaming you." Molly clapped her hands. "Deal. Tomorrow, we’re the most sinister clowns this neighborhood’s ever seen."
created_at
"2026-02-16T21:26:01.445244+00:00"
2/16/2026, 9:26:01 PM
model
"deepseek/deepseek-chat-v3"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
449
moderation_latency
(null)
generation_time
10731
tokens_prompt
77
tokens_completion
352
native_tokens_prompt
66
native_tokens_completion
335
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
0.00031927
router
(null)
provider_responses
0
id
"chatcmpl-RYmeGy4DwDzXqMseH7efIxez"
status
200
is_byok
false
latency
50
endpoint_id
"5294d55f-9012-496b-8f22-8cc919432dcd"
provider_name
"DeepInfra"
model_permaslug
"deepseek/deepseek-chat-v3"
api_type
"completions"
id
"gen-1771277150-1wqMG3jD73cKWtqR2vea"
upstream_id
"chatcmpl-RYmeGy4DwDzXqMseH7efIxez"
total_cost
0.00031927
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
99.7443%
Matches word count
n/a
neededClean
false
words
204
99.4338%
Dialogue to Total Word Ratio
Ratio: 87.25%, Deviation: 2.75%
neededClean
false
wordsTotal
204
wordsDialogue
178
99.5890%