NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY7JKH4G73NAHG4NN958F7D
Run Details
Model:
Gemma 3 12B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
17.0s
Cost:
$0.000095
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.01%
Dialogue to Total Word Ratio
(details)
0.0033%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 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
The aroma of stale coffee and desperation hung heavy in Molly’s cramped apartment. A single, bare bulb illuminated a table littered with crumpled napkins, a half-eaten bag of cheese puffs, and a frankly alarming number of child-sized drawings. Bill, a man whose optimism was inversely proportional to his success rate, was gesturing wildly with a slightly melted lollipop. “Okay, so picture this, Molly. A veritable cornucopia of joy! Wrapping paper, ribbons, miniature ponies…all ours!” Molly pinched the bridge of her nose. “Bill, we’re planning to rob a children’s birthday party. A *birthday* party. Not Fort Knox." “Details, details! The principles are the same! Strategic infiltration, precision extraction, clean getaway. Think Ocean’s Eleven, but with more balloons.” He popped the lollipop in his mouth, leaving a sticky trail. “And considerably less George Clooney.” “Ocean's Eleven didn't involve dodging sugar-fueled preschoolers with questionable coordination.” Molly slid a thumbnail across a sketch of a particularly aggressive-looking teddy bear. “What’s the target acquisition procedure here? Do we just waltz in and start grabbing presents?” “Absolutely not! We’re professionals, darling. We’re… strategic gift redistributors. I've been observing the target, Bartholomew Higgins the Third's fifth birthday. Significant security lapse. The bouncy castle.” Molly raised an eyebrow. "The bouncy castle is a security lapse? How?” “Distraction, my dear! Utter, glorious distraction! While the parents are busy policing rogue inflatable bouncers, we swoop in! Like shadows! Gift-snatching shadows!” Bill made a dramatic swooping motion that nearly knocked over a stack of comic books. “And how do we, ‘gift-snatching shadows,’ get past the, you know, *adults*?” Molly asked, her voice laced with skepticism. “Camouflage. Think about it. We dress as… clowns!” Bill beamed, clearly delighted with his own brilliance. Molly stared at him. “You want us to dress as clowns and rob a children’s birthday party? Bill, that’s… spectacularly stupid, even for you." “Hey! Clowns are familiar! Friendly! Plus, a little slapstick can disarm suspicion. Who’s going to suspect a clown of grand larceny? It's genius, I tell you! Pure, unadulterated genius!” "It's going to attract more attention than a gorilla riding a unicycle." Molly retorted, picking up a cheese puff and examining it with a deep sense of disappointment. "What about the presents themselves? We can’t exactly stuff twenty Lego sets down our clown pants." "I've thought of that!" Bill declared triumphantly, pulling out a large, canvas shopping bag. "We utilize the mobile gifting unit! Discreet! Spacious! Slightly suspicious looking, but easily explained away as... well, a shopping bag!" Molly sighed, massaging her temples. “And what’s your plan if we get caught?” "Deny everything! Blame the bouncy castle. Say it made us do it. Maybe juggle some rubber chickens for the police?” Bill offered, utterly serious. “Bill,” Molly said slowly, “I'm going to need a *lot* more coffee.”
created_at
"2026-02-20T19:14:10.608Z"
2/20/2026, 7:14:10 PM
model
"google/gemma-3-12b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
440
moderation_latency
(null)
generation_time
16944
tokens_prompt
93
tokens_completion
833
native_tokens_prompt
75
native_tokens_completion
708
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
0.00009504
router
(null)
provider_responses
0
endpoint_id
"eb06dc92-5a16-47ec-a776-6ef956457c47"
id
"chatcmpl-RkoZ558SBBAwGeoDSe2do5jK"
is_byok
false
latency
63
model_permaslug
"google/gemma-3-12b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771614850-sGqUDMJOoXgufbR46JhV"
upstream_id
"chatcmpl-RkoZ558SBBAwGeoDSe2do5jK"
total_cost
0.00009504
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
465
0.01%
Dialogue to Total Word Ratio
Ratio: 63.87%, Deviation: 13.87%
neededClean
false
wordsTotal
465
wordsDialogue
297
0.0033%