NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY830KZ1X9QJE3RAEQEK1R1
Run Details
Model:
Ministral 3 8B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
3.3s
Cost:
$0.000101
Provider:
Mistral
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 neon glow of a flickering "Happy Birthday!" balloon cast long shadows across the backseat of Bill’s beat-up van as Molly adjusted the strap of her backpack, her fingers tapping an impatient rhythm against the leather. The party was in full swing inside the house—laughter, squeals, and the unmistakable clink of plastic cups filled with juice. Perfect. Bill leaned back in his seat, arms crossed, a smirk playing on his lips. "You sure about this, Molly? Stealing from kids? That’s like robbing a library—where’s the fun in that?" Molly shot him a look that could curdle milk. "First of all, they’re not *kids*, they’re *consumers*. And second, you’ve been saying that since we scoped out the place. The parents are the real targets—they’re the ones who’ll notice if we’re not there to *deliver* the presents." Bill chuckled, shaking his head. "Oh, so now we’re *delivery guys* who got sidetracked? That’s a new one." "Bill, focus. The security’s a joke—just a motion-sensor light and a dog that’s probably more interested in sniffing its own butt than us. And the parents? They’re all too busy making sure their little angels don’t eat the cake raw." Bill held up a hand. "Wait, wait—you’re telling me we’re breaking into a *children’s party* because the parents are too distracted to notice us? That’s like robbing a bank during naptime." Molly rolled her eyes. "It’s *strategy*, Bill. And the best part? We’re not just stealing the presents—we’re *replacing* them with something *better*. Think about it: a kid unboxes a brand-new toy, only to find out it’s a *real* diamond-encrusted iPad. Their parents will be *thrilled*." Bill whistled low. "Now that’s a heist with *legacy*." Molly grinned. "Exactly. And by the time they figure out what happened, the presents will be long gone—sold to some poor sap who thinks he’s getting a steal at a garage sale." Bill leaned forward, suddenly serious. "You know, for a plan that involves stealing from kids, I’m weirdly proud of you." Molly smirked. "That’s because you’re a softie at heart, Bill. Now, let’s go before the cake gets eaten." Bill sighed, reaching for the door handle. "Fine. But if we get caught, I’m blaming you for *encouraging* me to steal from children." Molly laughed as she grabbed her bag. "Deal. And Bill?" "Yeah?" "Don’t forget the duct tape." Bill groaned. "I *hate* you."
created_at
"2026-02-20T19:23:08.3Z"
2/20/2026, 7:23:08 PM
model
"mistralai/ministral-8b-2512"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
238
moderation_latency
(null)
generation_time
3037
tokens_prompt
77
tokens_completion
642
native_tokens_prompt
69
native_tokens_completion
607
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
0.0001014
router
(null)
provider_responses
0
endpoint_id
"c85d26e1-1e0c-449b-9775-2afce7ae510b"
id
"fd8ff522200e4292aa5b5d2b0501100d"
is_byok
false
latency
238
model_permaslug
"mistralai/ministral-8b-2512"
provider_name
"Mistral"
status
200
api_type
"completions"
id
"gen-1771615388-iAZ5jIftIKkgCi6n7foC"
upstream_id
"fd8ff522200e4292aa5b5d2b0501100d"
total_cost
0.0001014
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
403
0.00%
Dialogue to Total Word Ratio
Ratio: 69.73%, Deviation: 19.73%
neededClean
false
wordsTotal
403
wordsDialogue
281
0.0000%