NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HJ42X8KT9TKFJCJF5RMV85
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
25.9s
Cost:
$0.030054
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
60.28%
Matches word count
(details)
95.95%
Dialogue to Total Word Ratio
(details)
78.1136%
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
The rusted interior of the surveillance van smelled intensely of mildew, stale french fries, and high-stakes criminal ambition. Molly peeled back the floral privacy curtain of the rear window, adjusting a pair of heavy brass binoculars. Across the cul-de-sac, a suburban fortress of pastel perfection stood guarded by spinning rainbow pinwheels and a yellow bouncy castle that wheezed rhythmically in the humid afternoon air. The gift table sat in the center of the patio like an unguarded treasure galleon. "Look at that perimeter," Molly murmured. "Sticky fingers at every single entrance." Behind her, Bill groaned from his perch atop an upturned plastic milk crate. He was currently wedged into an oversized, lime-green plush dinosaur costume, his graying beard protruding awkwardly from the felt maw of a smiling Tyrannosaurus rex. Sweat beaded across his forehead as he tugged at the stubborn seam. "Remind me why we are hitting a seven-year-old girl’s birthday party instead of a bank?" Bill asked. Between the front seats hung a hand-drawn blueprint of the target zone, rendered meticulously in stolen restaurant crayons. Molly had marked every major defensive perimeter with miniature skull stickers, reserving a glittering pink unicorn symbol for the terrifying president of the neighborhood association. "Because banks do not stock limited-edition robotic ponies," Molly replied smoothly, tapping a red marker against the drawing. "That patio table holds forty grand in mint-condition luxury loot." She pointed with surgical precision toward a towering stack of metallic foil packages topped with giant satin bows. Bill grunted, straining his arms behind his back to free the bulky costume fabric. "My lizard tail is caught in the zipper again," he muttered, twisting around before peering over her shoulder. "What about the hostiles?" "Eighteen second-graders hopped up on blue frosting," Molly said, tracking a pack of squealing children charging past the sprinkler system. "They have zero tactical discipline and no fear of God." "And the clown?" Bill asked. "Barnaby the Balloon Guy," Molly whispered, pointing her binoculars toward a hulking figure wearing red overalls. "Dishonorably discharged from the county circus circuit. Cross him, and he squeaks you into submission." Bill winced, eyeing the performer's suspiciously muscular forearms. The stakes were high, but the window of opportunity was rapidly approaching. Molly checked the stopwatch hanging around her neck, then reached into her duffel bag to arm herself with a tactical water gun loaded with sticky cherry syrup. "And what is our distraction?" Bill asked, pulling the dinosaur hood over his ears. "The pinata is rigged with glitter and raw eggs," Molly grinned coldly. "Total suburban anarchy will erupt in five minutes." "Shovel the loot into your dinosaur pouch and retreat to the compost bin," she instructed, sliding open the van door an inch. "If a soccer mom tackles me, I am squealing on you," Bill said. "Quiet, Bill. Put your game face on." Molly pushed the side door open, and the duo crept into the shrubbery.
created_at
"2026-09-02T17:19:30.735Z"
9/2/2026, 5:19:30 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1346
moderation_latency
(null)
generation_time
25838
tokens_prompt
75
tokens_completion
2835
native_tokens_prompt
67
native_tokens_completion
8001
native_tokens_completion_images
0
native_tokens_reasoning
7356
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
0.030054
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369570-fgq6uSCZraRwg8l4yd26"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369570-Xrvlozvjx0HNw1QqAz12"
upstream_id
"olqYaofSMI3W_uMPx7qoKQ"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"olqYaofSMI3W_uMPx7qoKQ"
is_byok
false
latency
1346
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.030054
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
60.28%
Matches word count
n/a
neededClean
false
words
485
95.95%
Dialogue to Total Word Ratio
Ratio: 30.91%, Deviation: 0.91%
neededClean
false
wordsTotal
495
wordsDialogue
153
78.1136%