NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM63DSQR153C6CFBX3JG0FK
Run Details
Model:
DeepSeek V3 (2024-12-26)
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
12.7s
Cost:
$0.000455
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
Molly adjusted her fake mustache in the rearview mirror while Bill fiddled with a party hat two sizes too small. "You sure this is the right move?" Bill asked, squashing the hat onto his head. "Stealing from kids feels... icky." Molly scoffed. "They’re *rich* kids, Bill. They’ll get new toys before the tears even dry. Plus, have you seen the size of those gift bags? We’re talking top-tier loot." Bill sighed. "Fine. But if we get caught, we’re telling the cops we’re clowns. Professional ones." "Deal." Molly pulled out a map of the backyard. "Okay, here’s the plan: You distract the parents with your ‘magic tricks’—" "My *what*?" "—while I load the presents into the piñata-shaped getaway van." Bill blinked. "We have a piñata van?" "We *will* in ten minutes. I bribed a guy." A beat. "You’re terrifying." Molly grinned. "Stick with me, Bill. By tonight, we’ll be knee-deep in finger paints and disappointment." Bill groaned. "I *hate* finger paints." "Then grab the confetti bombs and let’s go. Time’s wasting, and the cake’s calling my name." Bill shook his head. "Remind me why I partner with you?" "Because no one else tolerates your whining." "Fair."
created_at
"2026-02-16T21:36:13.439721+00:00"
2/16/2026, 9:36:13 PM
model
"deepseek/deepseek-chat-v3"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1248
moderation_latency
(null)
generation_time
12707
tokens_prompt
77
tokens_completion
337
native_tokens_prompt
66
native_tokens_completion
330
native_tokens_completion_images
0
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
0.0004554
router
(null)
provider_responses
0
id
"f70f227ad06e14f25fbcf18857c26ea2"
status
200
is_byok
false
latency
1248
endpoint_id
"cefb5a1e-9fea-4496-9a64-4e1b11a7cc8c"
provider_name
"Novita"
model_permaslug
"deepseek/deepseek-chat-v3"
api_type
"completions"
id
"gen-1771277760-l5B5GU3g5HRpTdnPvBX7"
upstream_id
"f70f227ad06e14f25fbcf18857c26ea2"
total_cost
0.0004554
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Novita"
Evaluation details
Result
Evaluator
Details
Meta Data
99.9990%
Matches word count
n/a
neededClean
false
words
199
0.0012%
Dialogue to Total Word Ratio
Ratio: 68.34%, Deviation: 18.34%
neededClean
false
wordsTotal
199
wordsDialogue
136
50.0001%