NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HJ35NTM69TDFXND8CHDM9D
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
32.6s
Cost:
$0.042380
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
99.74%
Matches word count
(details)
99.15%
Dialogue to Total Word Ratio
(details)
99.4451%
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 slammed the master operational blueprint onto the hood of their battered surveillance van. Technically, it was a greasy paper placemat from a local pancake franchise, but her crimson crayon markings were executed with the chilling precision of a seasoned mercenary. She tapped a wobbly circle drawn directly over the suburban patio, her tactical flashlight beam trembling slightly against the heavy layers of wax. "The primary target zone," Molly whispered grimly. "Fourteen hundred hours. Maximum sugar, minimum adult supervision." Bill leaned heavily against the dented radiator, adjusting a reinforced tool belt that clinked with an embarrassing assortment of tension wrenches, heavy wire cutters, and three punctured strawberry juice boxes. He rubbed his stubbled chin, squinting down through the gloom at the crude sketch with the profound skepticism of a veteran thief who possessed standards. "It’s a bouncy castle, Molly. You’ve actually drawn sniper perches on an inflatable pirate ship." Molly shoved a sticky grape lollipop aside with the tip of her penknife. Her jaw was set like granite. She unrolled a crumpled strip of photographs torn from a neighborhood newsletter, affixing them to the cracked windshield with a sticky wad of blue poster putty to emphasize their tactical significance. "That castle is an elevated observation deck occupied by hyperactive sentries with sticky hands. Our bounty rests on the trestle table beside the cupcake pyramid. Exactly thirty wrapped boxes. Expensive toy sets and two motorized drones." Bill sighed, cross-referencing her hasty crayon arrows with a cracked silver stopwatch he had lifted from a middle school gym teacher three years earlier. He rubbed his aching temples, already feeling the sharp sting of moral decay creeping into his chest. "Are we really sinking to robbing a toddler? The cops will not need dogs; they will just follow the glitter." Molly began pacing the narrow, rust-eaten aisle of the cargo hold, nimbly sidestepping two empty bags of barbecue chips and a pair of customized burlap sacks. She tested the nylon cord of her oversized duffel bag with a violent snap, entirely deaf to her partner’s sudden attack of conscience. "Glitter is an acceptable operational casualty. We strike during the cake cutting. The singing provides acoustic camouflage, and the sparklers will blind the perimeter for eight critical seconds." Bill turned slowly toward the dark corner of the van and gingerly hoisted their disguise by two fingers. He held up a voluminous, aggressively white jumpsuit smothered in hundreds of hot-glued cotton balls, staring at the synthetic abomination with a gaze that broadcast pure, unadulterated misery to the world. "And this? You promised discreet clothes. I look like an exploded mattress pile." Molly pulled a pair of mirrored sunglasses over her eyes without blinking a single time. "You are Barnaby the Gentle Cloud. Clowns draw suspicion, Bill; clouds bring serenity. Zip up your precipitation and grab the sacks. Those toys are ours tonight." Bill squeezed his weary eyes shut, quietly mourned the dignified safecracker he used to be, and shoved his trembling combat boots straight down into the fluffy white cotton.
created_at
"2026-09-02T17:19:00.803Z"
9/2/2026, 5:19:00 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1554
moderation_latency
(null)
generation_time
32592
tokens_prompt
75
tokens_completion
3576
native_tokens_prompt
67
native_tokens_completion
11288
native_tokens_completion_images
0
native_tokens_reasoning
10647
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
0.04238025
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369540-hUQmlwoUIDhaV5LthzyR"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369540-7bSvPgA2JlhA1LexxBRu"
upstream_id
"hFqYaq-DNf6W1MkPm8_MwQc"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"hFqYaq-DNf6W1MkPm8_MwQc"
is_byok
false
latency
1554
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.04238025
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
99.74%
Matches word count
n/a
neededClean
false
words
504
99.15%
Dialogue to Total Word Ratio
Ratio: 29.59%, Deviation: 0.41%
neededClean
false
wordsTotal
507
wordsDialogue
150
99.4451%