NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHXPC724MKWFSE7H3Y4CHYAP
Run Details
Model:
Mistral Small 3.2 24B
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
6.7s
Cost:
$0.000093
Provider:
Mistral
Evaluation summary
Score
Evaluator
Details Link
14.30%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
7.1507%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 oversized sunglasses and squinted at the brightly colored banner flapping in the breeze. "This is it," she whispered, nudging Bill with her elbow. "The Great Uncle Larry's 7th Birthday Extravaganza. A goldmine of unsupervised presents." Bill scratched his chin, eyeing the inflatable castle bouncy castle. "You sure this is legal? I mean, stealing from kids?" Molly scoffed. "Legal? Bill, we’re thieves. Our moral compass is a compass that only points to ‘shiny things.’ Besides, kids don’t even know what money is. They just cry until someone gives them candy." Bill sighed. "Fair point. So, what’s the plan?" Molly grinned. "Simple. You distract the parents with your ‘world-famous’ magic tricks, and I’ll sneak in through the back. Kids love magic. They’ll be too busy oohing and aahing to notice me stuffing their loot into a duffel bag." Bill frowned. "My magic tricks are *not* world-famous." Molly patted his shoulder. "Exactly. That’s why it’s perfect. No one will suspect a thing." Bill groaned. "This is a terrible idea." Molly winked. "That’s why we’re going to do it."
created_at
"2026-02-20T14:13:35.438Z"
2/20/2026, 2:13:35 PM
model
"mistralai/mistral-small-3.2-24b-instruct-2506"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
263
moderation_latency
(null)
generation_time
6566
tokens_prompt
77
tokens_completion
299
native_tokens_prompt
69
native_tokens_completion
288
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
0.0000933
router
(null)
provider_responses
0
endpoint_id
"23fd6b91-74d8-45fc-ac00-cf141ec7f4a5"
id
"2e2c95c65cb54e5c968b76868aafeb69"
is_byok
false
latency
263
model_permaslug
"mistralai/mistral-small-3.2-24b-instruct-2506"
provider_name
"Mistral"
status
200
api_type
"completions"
id
"gen-1771596815-aAYaZSzEVP0GC0cIeaJx"
upstream_id
"2e2c95c65cb54e5c968b76868aafeb69"
total_cost
0.0000933
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mistral"
Evaluation details
Result
Evaluator
Details
Meta Data
14.30%
Matches word count
n/a
neededClean
false
words
179
0.00%
Dialogue to Total Word Ratio
Ratio: 72.63%, Deviation: 62.63%
neededClean
false
wordsTotal
179
wordsDialogue
130
7.1507%