NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HJCN2TERA3Y4XHGA21KR0K
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
37.2s
Cost:
$0.047259
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
86.38%
Matches word count
(details)
74.37%
Dialogue to Total Word Ratio
(details)
80.3763%
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 smoothed the crinkled map over the steering wheel of their surveillance van. Across the suburban street, cheerful accordion music echoed from the Higgins backyard. "Look at the blueprints, Bill. This is the big leagues." "Molly, that’s a greasy cocktail napkin with crayon on it." "It’s a tactical schematic, you uncultured brute. Target time is fourteen-hundred hours. We breach through the hydrangeas, evade the bouncy castle, and secure the gift table." Bill adjusted his dark aviator sunglasses, squinting through the grease-stained windshield at the cluster of pastel balloons. "The bouncy castle has a perimeter?" "It’s guarded by six sugar-crazed second-graders. You step on that vinyl without socks, you’re dead." "I brought non-slip grip socks from the trampoline park." Molly nodded with grim approval, tapping her ballpoint pen against the crudely drawn layout of the crowded backyard patio. "Brenda Higgins runs a tight ship. Eyes everywhere, mostly checking labels for gluten. If Brenda spots us, we deploy the primary decoy." "Which is?" "An artisanal baked brie left unattended on the kitchen island. It’ll buy us four minutes." "Brilliant. And the father?" "Greg’s at the grill, four beers deep. Neutralized." She pointed a gloved finger toward a tall, ominous figure twisting long balloons beside the patio umbrella. "What about the hired muscle?" "Bangles the Clown. Shifty eyes. Intel says he’s armed with balloon animals, but I suspect heavy-duty glitter and high-velocity confetti. If you hear squeaking, drop to the grass." "Right. What’s the haul?" "A motorized mini-Jeep, two drones, and an envelope of cash from Grandma Higgins." Bill shifted uncomfortably in the cramped passenger seat, his knees pressing hard against the dented glove compartment. "A motorized Jeep? How are we getting that over the cedar fence?" "You’re driving it out the side gate." "Me? I’m six-foot-two and two hundred pounds! My knees will be jammed straight into my chin! I’ll look like an overgrown toddler in a plastic Barbie ride!" "Then pedal fast if the battery dies. I’ll follow behind with the video games and the Lego sets." Outside, a sudden chorus of high-pitched shrieks erupted across the lawn as someone suspended a papier-mâché donkey. "What if the children swarm us, Molly? You’ve never seen them when the juice boxes run dry. They turn feral. Last Thanksgiving my nephew tackled an uncle over a fruit roll-up." "That’s why we strike during the main event." "The cake cutting?" "The piñata. The moment the blindfolded kid swings the bat, chaos reigns. The parents film, the kids dive for candy, and nobody notices two adults hauling five hundred dollars of high-end plastic down the driveway." "It’s ruthless, Molly. He’s turning seven." She glared back at him, pulling a bright pink ski mask over her brown hair. "Timmy doesn't appreciate mint-condition packaging, Bill. He tears through cellophane like a rabid raccoon. We’re doing the collector community a genuine service, and that Lego Millennium Falcon belongs in an adult’s display cabinet." "Fair point. Hand me my clown nose. Let’s go make some money." Bill snapped the red sponge ball onto his nose and quietly clicked open his door.
created_at
"2026-09-02T17:24:11.491Z"
9/2/2026, 5:24:11 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3474
moderation_latency
(null)
generation_time
37109
tokens_prompt
75
tokens_completion
4007
native_tokens_prompt
67
native_tokens_completion
12589
native_tokens_completion_images
0
native_tokens_reasoning
11825
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
0.047259
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788369851-EMz6Vh0OtMufv0BII0gL"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788369851-d0JSmLR62Vv9I951zOCG"
upstream_id
"u1uYaqHBIoz2jMcPr7DHyQ8"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"u1uYaqHBIoz2jMcPr7DHyQ8"
is_byok
false
latency
3474
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.047259
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
86.38%
Matches word count
n/a
neededClean
false
words
511
74.37%
Dialogue to Total Word Ratio
Ratio: 72.43%, Deviation: 2.43%
neededClean
false
wordsTotal
526
wordsDialogue
381
80.3763%