NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-19-01KHVPM75Q90DMSX72RGCN7WC7
Run Details
Model:
Gemini 3.1 Pro (Preview)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 55s
Cost:
$0.120350
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
100.0000%
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
Molly adjusted the heavy brass binoculars against her face, scanning the treacherous terrain of the suburban backyard. It was a fortress of brightly colored plastic. "What is the precise operational status of the primary target zone today, Bill? Are the sentries still actively patrolling the perimeter of the trampoline area?" Bill shifted his weight in the passenger seat of their surveillance vehicle. He was closely reviewing a detailed map scrawled entirely onto a cocktail napkin. "The perimeter is currently secured by a towering inflatable moat. The civilians refer to it as a bouncy house, but it is a tactical nightmare." Molly let out a long, frustrated breath and slowly lowered her lenses. She massaged the bridge of her nose to ward off an impending migraine. "And what about the designated payload? Have you visually confirmed the location of the gifts? We cannot afford to make a mistake during the extraction." Bill reached into a crumbled bag of stale tortilla chips. He crunched one rather loudly before snapping his fingers and pointing toward the wooden patio. "The loot is clustered together on the patio table. The goods are entirely concealed beneath layers of thick dinosaur paper. Thirty individual high value units." Molly tapped her fingers against the cracked leather of the steering wheel. She mentally calculated the optimal angles while completely ignoring the nearby mariachi music. "We need a massive distraction immediately. We must deploy something incredibly loud to pull those hyperactive civilian guards entirely away from the unguarded patio area." Bill smiled, happily revealing a slightly chipped front tooth. He reached downward and firmly patted a rusty metal lever situated right next to his hip. "I genuinely believe that I should release our captive clown right now. He will certainly cause enough chaos and confusion to clear the entire lawn." Molly slowly turned her head. She fixed him with a cold, unblinking glare that had once famously caused a museum curator to burst into tears. "We do not currently possess a clown for this particular mission, Bill. What on earth are you talking about? There are zero clowns authorized today." A muffled thumping sound suddenly echoed off the vehicle frame. The trunk shook ever so slightly under the heavy weight of someone angrily kicking it. "I hired one this morning. He is waiting in the trunk. His professional stage name is Giggles, and he gets angry when he is confined." Molly squeezed her eyes firmly shut. A prestigious career of stealing priceless diamond tiaras had somehow degraded into aggressively robbing a tiny neighborhood birthday party. "You locked an actual working birthday clown inside the hot trunk of our designated getaway vehicle? Are you completely out of your mind right now?" Bill nodded proudly while adjusting his dark ski mask. He grabbed two reinforced canvas sacks from the backseat and prepared to breach the wooden fence. "It is a standard diversion tactic. Once I pop the latch, he will run wildly. Grab the Lego sets and never look back at him."
created_at
"2026-02-19T19:39:28.655Z"
2/19/2026, 7:39:28 PM
model
"google/gemini-3.1-pro-preview-20260219"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3464
moderation_latency
(null)
generation_time
114884
tokens_prompt
77
tokens_completion
3658
native_tokens_prompt
67
native_tokens_completion
10018
native_tokens_completion_images
0
native_tokens_reasoning
9417
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
0.12035
router
(null)
provider_responses
0
endpoint_id
"2abf5677-0826-45d9-8470-3d62c31d54a1"
is_byok
false
latency
3464
model_permaslug
"google/gemini-3.1-pro-preview-20260219"
provider_name
"Google AI Studio"
status
200
api_type
"completions"
id
"gen-1771529968-E5NPWzXbckVbK6Kw5RGP"
upstream_id
(null)
total_cost
0.12035
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
100.00%
Dialogue to Total Word Ratio
Ratio: 50.00%, Deviation: 0.00%
neededClean
false
wordsTotal
500
wordsDialogue
250
100.0000%