NC Bench
Overview
About
Tests
Models
Model
Gemini 3 Flash (Preview)
Test
Dialogue tags
Scenario
Write 500 words with 70% dialogue
Run ID
2026-02-08-01KGYWZ7N5AHNDCA8XCBBY39A6
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
"Check the perimeter again," Bill whispered, adjusting his overly tight tuxedo. "I see a bouncy castle. Is that a tactical obstacle or just a hazard?" Molly didn’t look up from her blueprint, which was actually a crayon drawing of the backyard she’d bribed off a disgruntled six-year-old. "It’s a distraction, Bill. Focus. The objective is the mahogany sideboard in the sunroom. That’s where the high-value loot is concentrated." "By loot, you mean the Lego Star Wars sets and the remote-controlled drones?" "I mean the liquid assets, Bill. Do you have any idea what the resale value on a mint-condition Galactic Destroyer is these days?" "I’m more worried about the biological warfare," Bill said, pointing a trembling finger toward the sliding glass door. "Look at them, Molly. They’re covered in frosting. They have sticky fingers. If one of them touches this suit, I’m compromised." "Grow a spine. Did you pack the suppression devices?" Bill patted his pockets. "I’ve got three bags of organic kale chips and a dozen sugar-free juice boxes. If they charge, I’ll create a bottleneck at the juice station." "Good. Now, the clown. What’s his status?" "Target name: Giggles. Real name: Arthur Pringle. He’s currently distracted by a balloon animal malfunction. We have a three-minute window before he starts the magic show." "That’s our hex," Molly said, snapping her gum. "When the kids are distracted by the disappearing pigeon, we slip through the doggy door." "The doggy door? Molly, I’ve been hitting the carbs. I won't fit." "You’ll fit if I grease you up with the artisanal butter from the refreshment table. Now, listen. Once we’re inside, we ignore the plushies. We go straight for the boxes that rattle. If it sounds like plastic bricks, it’s coming with us." "Wait," Bill hissed, gripping her arm. "Look at the birthday boy’s mother. She’s reached for a clipboard. She’s an organizer, Molly. A tactician." "She’s a suburbanite with a Pinterest addiction. We’ve outsmarted Interpol, Bill. We can handle a woman named Karen." "She’s wearing a headset! Why is she wearing a headset?" "It’s for the pony coordinator! Stay calm. Do you have the tactical distraction ready?" Bill reached into his vest and pulled out a small, plush tabby cat. "If I squeeze its tail, it meows in a frequency only toddlers can hear. It’ll cause a stampede toward the sandbox." "Perfect. On my mark. Three. Two. One. Execute." "Wait! Molly, look! They’re serving the cake!" "Abort! Abort the entry!" Molly yelled, pulling him back into the bushes. "Why? This is our chance!" "Are you insane? That’s a three-tier chocolate ganache. If we go in now, we’ll be trampled by twenty-five sugar-crazed primates. We wait for the crash. We strike during the nap-time lull." "Right. Professionalism above all else." "Exactly. Now pass me those binoculars. I want to see if that gift bag has a Nintendo Switch."
created_at
"2026-02-08T15:12:30.411557+00:00"
Sunday, February 8th, 2026 at 3:12:30 PM GMT+00:00
model
"google/gemini-3-flash-preview-20251217"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
654
moderation_latency
(null)
generation_time
6679
tokens_prompt
61
tokens_completion
683
native_tokens_prompt
67
native_tokens_completion
732
native_tokens_completion_images
0
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
0.0022295
router
(null)
provider_responses
0
status
200
is_byok
false
latency
654
endpoint_id
"6e82f652-2693-4b60-bbe0-7362d6e345f0"
provider_name
"Google AI Studio"
model_permaslug
"google/gemini-3-flash-preview-20251217"
api_type
"completions"
id
"gen-1770563542-8yhNClfnw06ilKXejZcf"
upstream_id
(null)
total_cost
0.0022295
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
Evaluation details
Result
Evaluator
Details
Meta Data
75.1556%
Matches word count
n/a
neededClean
false
words
487
0.0042%
Dialogue to Total Word Ratio
Ratio: 84.19%, Deviation: 14.19%
neededClean
false
wordsTotal
487
wordsDialogue
410
37.5799%