NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5RTHY159PG9SGR4D5PZ1J
Run Details
Model:
DeepSeek V3 (2024-12-26)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
30.1s
Cost:
$0.000779
Provider:
DeepInfra
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 back in her chair, twirling a pencil between her fingers. “So, let me get this straight. You want to rob a *children’s party*?” Bill shrugged, adjusting his baseball cap. “Not just any children’s party. The *Pettigrew* party. You know, the ones with the mansion and the golden retriever named Mr. Biscuits? Their kid’s turning six, and the gifts are gonna be insane.” Molly raised an eyebrow. “Insane, huh? What are we talking? Toy trucks? Coloring books?” “No, Molly!” Bill slapped the table. “We’re talking designer dolls, tiny electric cars, maybe even a mini jet ski. I heard the grandparents are loaded.” “A mini jet ski,” Molly repeated flatly. “For a six-year-old.” “Exactly!” Bill leaned in, his voice dropping to a conspiratorial whisper. “And here’s the kicker—the party’s at a *petting zoo*. The parents are too busy cooing over baby goats to notice us slipping into the gift tent.” Molly sighed, rubbing her temples. “Bill, this is the dumbest idea you’ve ever had. And that’s saying something, considering last month’s ‘heist’ at the sock factory.” “Hey, that was a solid plan,” Bill protested. “I just didn’t account for the sock-eating security dog.” “Which *ate* your getaway car tires,” Molly reminded him. “Minor setback,” Bill said, waving it off. “This is different. It’s foolproof. We dress up as clowns, blend in with the entertainment, and—” “Clowns?” Molly interrupted. “Bill, you’re terrified of clowns.” “Not all clowns!” Bill argued. “Just the ones with the creepy teeth and the balloon animals that look like they’re judging you.” Molly groaned. “This is a disaster waiting to happen. What happens when the kids realize we’re not real clowns? What happens when Mr. Biscuits smells your fear and starts barking?” “I’ll handle Mr. Biscuits,” Bill said confidently. “I’ve got peanut butter in my pocket.” “Of course you do,” Molly muttered. She leaned forward, locking eyes with him. “Bill, listen to me. Even if we pull this off—*if*—we’re stealing from children. Do you really want to be the guy who swipes a six-year-old’s birthday gifts?” Bill hesitated, scratching his chin. “Well, when you put it like that…” “Exactly,” Molly said. “It’s heartless. It’s low. It’s—” “Genius!” Bill interrupted, snapping his fingers. “Because who’s gonna suspect the clowns? Nobody! It’s the perfect cover.” Molly stared at him, incredulous. “Bill, I’m not doing this. I have a reputation to uphold.” “Reputation?” Bill scoffed. “You’re the one who got us banned from the mall after the ‘incident’ with the escalator and the inflatable pool toys.” “That was *art*,” Molly shot back. “And besides, I already told you—no kids’ parties.” Bill leaned back, crossing his arms. “Fine. But when I’m rolling around in my mini jet ski next week, don’t come crying to me.” Molly smirked. “Deal. But when Mr. Biscuits bites you, don’t come crying to me.” Bill grinned, pulling out a map of the petting zoo. “It’s gonna be epic, Molly. Just wait and see.” Molly shook her head, muttering under her breath. “This is gonna blow up in our faces.” “That’s the spirit!” Bill exclaimed, clapping her on the back. “Now, let’s practice our clown laughs. Ready? Ha-ha-ha!” Molly groaned, burying her face in her hands. “I hate my life.”
created_at
"2026-02-16T21:30:43.688368+00:00"
2/16/2026, 9:30:43 PM
model
"deepseek/deepseek-chat-v3"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
348
moderation_latency
(null)
generation_time
30051
tokens_prompt
77
tokens_completion
950
native_tokens_prompt
66
native_tokens_completion
852
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
0.0007794
router
(null)
provider_responses
0
id
"chatcmpl-RT31LqVtEQ364cn2XSO5AbNZ"
status
200
is_byok
false
latency
54
endpoint_id
"5294d55f-9012-496b-8f22-8cc919432dcd"
provider_name
"DeepInfra"
model_permaslug
"deepseek/deepseek-chat-v3"
api_type
"completions"
id
"gen-1771277412-Q93paDfsLvBLODDAunpJ"
upstream_id
"chatcmpl-RT31LqVtEQ364cn2XSO5AbNZ"
total_cost
0.0007794
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0000%
Matches word count
n/a
neededClean
false
words
542
27.3473%
Dialogue to Total Word Ratio
Ratio: 75.09%, Deviation: 5.09%
neededClean
false
wordsTotal
542
wordsDialogue
407
13.6737%