NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-11-11-01M75PMQ2Y6P80T8CNAQPEQV34
Run Details
Model:
Gemini 4 Carbon (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 4s
Cost:
$0.048004
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
92.01%
Dialogue to Total Word Ratio
(details)
96.0020%
Overall
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
Molly slammed a crumpled napkin onto the sticky diner booth, tapping a grease stain with dramatic intensity while glancing around for eavesdropping waitresses. "Listen closely, Bill, because we only get one shot at the Miller residence before the sugar crash hits." Bill squinted through his smudged glasses, adjusting his counterfeit Rolex with a skeptical frown. "Is that ketchup smear supposed to be the perimeter fence, or the bouncy castle?" "That is the bouncy castle, and it is our primary blind spot," Molly whispered fiercely, leaning across her untouched pancakes. "While the six-year-olds bounce their little brains out, nobody will be watching the gift table." "Six-year-olds?" Bill recoiled against the vinyl seat, looking genuinely horrified by the revelation. "You promised me a high-stakes score! We are robbing a first-grader?" "Not just any first-grader. Little Timmy Miller. His dad owns three dental practices, and his grandma buys affection with cash. Do you know the black-market value of an unopened Millennium Falcon Lego set? It is practically untraceable bullion! Plus, the rich aunts always bring crisp fifty-dollar bills tucked inside musical greeting cards." Bill rubbed his aching temples, letting out a long, defeated sigh. "Fine, I see the financial angle. What is our cover? Caterers? Landscapers?" Grinning wickedly, Molly pulled a bright red foam nose from her leather jacket pocket and bounced it in her palm. "Absolutely not," Bill said flatly, holding up both hands in surrender. "I draw the line at clowning. Balloon animals give me hives, Molly, and oversized shoes ruin my sprint time." "Relax, you baby, I am the clown. You are going in as 'Exhausted Uncle Gary.' Just wear cargo shorts, hold a lukewarm beer, and stare blankly into space whenever adults make eye contact. Nobody questions a miserable relative at a kid’s party." Bill considered the role, nodding slowly as he stroked his stubbled chin. "Miserable is my resting state. So I blend in by the grill. How do we extract the loot without triggering a tantrum alarm?" "Simple," Molly said, tracing an escape route past the salt shaker. "At two-thirty, they bring out the Paw Patrol cake. Every eye in that backyard will be locked onto those candles. While I distract stragglers by pretending to choke on a kazoo, you back the minivan to the gate and sweep the gifts into lawn bags." "Wait a minute," Bill interrupted, tapping the napkin anxiously. "What if they open presents before cake? Have you accounted for rogue parenting? One impatient toddler tears some wrapping paper early and our whole timeline implodes!" Molly scoffed, waving away his amateurish concern with a flick of her wrist. "Please. Jennifer Miller runs a lifestyle blog. Her itinerary is color-coded on Pinterest. Cake at two-thirty, presents at three, and by three-oh-five, we are swimming in plastic bricks and Toys-R-Us gift cards." Bill stared at the foam nose, then met Molly’s manic gaze. "You do realize Toys-R-Us went completely bankrupt years ago, right, Molly?" "Details, Bill! Focus on the big picture and go buy some cargo shorts!"
created_at
"2026-11-11T16:21:47.752Z"
11/11/2026, 4:21:47 PM
model
"google/gemini-4-carbon-20261110"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5642
moderation_latency
(null)
generation_time
64436
tokens_prompt
75
tokens_completion
2010
native_tokens_prompt
97
native_tokens_completion
4781
native_tokens_completion_images
0
native_tokens_reasoning
4087
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"STOP"
service_tier
"default"
usage
0.048004
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.2; linux; x64))"
http_referer
(null)
request_id
"req-1794414107-ZmYN18Sl0T0x4t3OPCzT"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1794414107-46osFbY33VZAG1qo5hWD"
upstream_id
"G5b0apD1N9PFxdIPpMiUyAY"
provider_responses
0
endpoint_id
"f0c9681d-e99b-4bcb-8031-752429da222d"
id
"G5b0apD1N9PFxdIPpMiUyAY"
is_byok
false
latency
5641
model_permaslug
"google/gemini-4-carbon-20261110"
provider_name
"Google AI Studio"
status
200
total_cost
0.048004
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
501
92.01%
Dialogue to Total Word Ratio
Ratio: 71.29%, Deviation: 1.29%
neededClean
false
wordsTotal
519
wordsDialogue
370
96.0020%