NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HJ86XNNGGK3Z320V2K0GWB
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
43.1s
Cost:
$0.058554
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.90%
Dialogue to Total Word Ratio
(details)
99.9507%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 flattened the chalk blueprints across the stained steering wheel of their rusted surveillance van. The humid summer air inside smelled of warm upholstery and ancient cheese crackers. Through the tinted windshield, suburban chaos raged across an emerald lawn where eighteen hyperactive children galloped around a giant inflatable monster. A tired father struggled with an overloaded barbecue grill in the distance, sending plumes of burnt hotdog smoke into the sky. Molly checked the stopwatch hanging from her rearview mirror, her eyes narrowed with predatory calculation. She tapped the dull pencil against the cardboard diagram. "Target is Timmy Higgins. He turns eight today, meaning prime loot." Beside her, Bill peered through a pair of cheap binoculars, his oversized red foam nose bumping against the glass. He crunched thoughtfully on a handful of stale corn chips, spraying yellow crumbs across the dashboard. "Are we talking video games or socks?" Molly sighed deeply, using her sleeve to wipe salty debris off the tactical perimeter map. Across the street, the sun reflected off a shimmering mountain of silver wrapping paper piled high on a white trestle table. "Drones, handheld consoles, and thick envelopes of cash from guilty grandparents." Bill gasped, wiping his buttery fingers on his spotted overalls. "Timmy has an uncle in private equity." Molly nodded gravely. "Phase one starts at two o'clock sharp when the cake appears. While parents document the sugar crash, we breach the patio slider." Bill lowered the binoculars and scratched his neon wig with a squeaky rubber mallet. "Does the bouncy castle block our escape route?" Molly grabbed the mallet from his grip and threw it into the backseat. "You are not bouncing anywhere. You enter as Sparkles the Clown, bag the haul in contractor sacks, and egress through the hedge." Bill adjusted his floppy silk necktie, looking genuinely disappointed. "Do I have to perform any magic?" Molly leveled a cold, merciless stare at him. "No sleight of hand. Just straight theft." Bill slumped against the passenger door with a heavy sigh. "Pity. I practiced pulling a dead pigeon from this hat all morning." She pinched the bridge of her nose, fighting an oncoming headache. He was reckless, deeply eccentric, and smelled faintly of vinegar, but nobody could haul eighty pounds of wrapped merchandise over a cedar fence faster than him. A career spent robbing banks had never prepared her for the sheer logistical terror of an affluent suburban birthday party. She reached into the footwell and lifted a heavy duffel bag, checking its contents one final time before the countdown began. "Inventory check. Tape? Smoke grenades?" Bill rummaged through the giant pockets of his clown pants, pulling out items with theatrical flair. "Juice boxes, string cheese, and four scented smoke bombs." She groaned, checked her watch, and slid a dark mask over her face. "Remember the golden rule. Leave no ribbon behind, and never twist a balloon animal." Bill grinned, revealing a gold tooth, and opened the van door. "Let us go ruin this spoiled boy's sweet afternoon."
created_at
"2026-09-02T17:21:45.917Z"
9/2/2026, 5:21:45 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1397
moderation_latency
(null)
generation_time
43096
tokens_prompt
75
tokens_completion
4589
native_tokens_prompt
67
native_tokens_completion
15601
native_tokens_completion_images
0
native_tokens_reasoning
14955
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
0.058554
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369705-4nX8rB2Esb54ciUgEp2r"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369705-lNl6dJ7SVKnFYZ9x79ln"
upstream_id
"KVuYaoODPYuq_uMP8oe9uQ0"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"KVuYaoODPYuq_uMP8oe9uQ0"
is_byok
false
latency
1397
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.058554
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
99.90%
Dialogue to Total Word Ratio
Ratio: 30.14%, Deviation: 0.14%
neededClean
false
wordsTotal
501
wordsDialogue
151
99.9507%