NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HH80D4TCVNY7Q9368WB09R
Run Details
Model:
Gemini 3.8 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
23.1s
Cost:
$0.029979
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
60.28%
Matches word count
(details)
99.17%
Dialogue to Total Word Ratio
(details)
79.7228%
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
"Look at the blueprints, Bill." Molly slammed a crumpled sheet of construction paper onto the dashboard of their battered surveillance van. Bill leaned over the steering wheel, squinting through the grease-stained windshield. "Molly, this is drawn in purple glitter crayon." "Purple represents maximum tactical priority. Now, pay attention. Target is the Jenkins residence. Suburban bi-level, heavy perimeter defenses." Molly smoothed the paper flat against the vinyl dash. "By defenses, you mean a rotating garden sprinkler?" Bill asked, drumming his fingers against the worn steering column. "A motion-activated oscillating hazard, Bill! Never underestimate suburban paranoia." Molly tapped the waxy purple sketch with an unclipped fingernail. "At fourteen-hundred hours sharp, the payload arrives." "You mean the bouncy castle rental?" "I mean an inflatable Trojan horse. While the hostiles are busy suffocating each other inside the nylon thunderdome, we breach through the basement laundry chute." Bill pinched the bridge of his nose, exhaling a long, weary sigh. "Molly, they are seven-year-olds. They aren’t hostiles. They’re wearing paper crowns and eating dinosaur nuggets." "Dinosaur nuggets soaked in adrenaline and pure sugar! If they catch us, they will swarm like piranhas. Have you seen little Timmy Jenkins? The kid bites." "He lost his front teeth on Tuesday." "Which makes his jaw dangerously unpredictable." Molly reached into the footwell and hauled up a heavy black duffel bag, unzipping it with a sharp jerk. "Check your inventory." Bill peered inside, fishing out a miniature carton. "A box of juice boxes and three party noisemakers? Molly, what is this?" "The apple juice is laced with crushed melatonin. It neutralizes the sentry." "Barnaby the golden retriever?" "Barnaby is a gluttonous monster and our greatest tactical threat." "I thought the clown was our greatest threat." Bill checked his watch, his knuckles white against the wheel. "Bongo the Clown is an amateur. When he starts twisting balloon poodles, that is our window. We have ninety seconds." "Ninety seconds to do what?" "To sweep the dining room gift table. All of it. The Lego sets, the drone, the boxed console." "Are you sure there's a console?" "Grandma Jenkins drives a Lincoln, Bill. That silver box with the giant velvet bow isn't socks. We bag the haul, dump it into the wheelbarrow, and slip through the side gate." Molly leaned in close, her eyes gleaming with manic ambition. "And if Timmy's mother intercepts us in the kitchen?" "You trigger the primary diversion." "Which is?" "You scream that the ice cream cake is melting on the counter. The entire domestic structure will descend into immediate hysteria." Bill paused, letting the cold logic settle over him. "Terrifyingly plausible." "I’ve run the simulations. Now take your disguise." Molly tossed two cardboard cones onto his lap. Bill held one up by its thin elastic string, staring at the cartoon puppy in disbelief. "Are these Paw Patrol party hats?" "Put it on, Bill. Marshall the Dalmatian walks away with the crown jewels today."
created_at
"2026-09-02T17:04:10.667Z"
9/2/2026, 5:04:10 PM
model
"google/gemini-3.8-flash-20260902"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1736
moderation_latency
(null)
generation_time
23042
tokens_prompt
75
tokens_completion
2891
native_tokens_prompt
67
native_tokens_completion
7981
native_tokens_completion_images
0
native_tokens_reasoning
7285
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
0.029979
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788368650-uEoHYD7GMqPQIanY7jNX"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788368650-M7LTK1urwomr0WOoDd0g"
upstream_id
"CleYatnGLI7SjMcPxsSb4AQ"
provider_responses
0
endpoint_id
"d376f781-516d-42f9-936d-46f19abca6cc"
id
"CleYatnGLI7SjMcPxsSb4AQ"
is_byok
false
latency
1736
model_permaslug
"google/gemini-3.8-flash-20260902"
provider_name
"Google AI Studio"
status
200
total_cost
0.029979
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
99.17%
Dialogue to Total Word Ratio
Ratio: 69.59%, Deviation: 0.41%
neededClean
false
wordsTotal
490
wordsDialogue
341
79.7228%