NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCTVPEEEMW1B4556PW6C5SQ
Run Details
Model:
Gemini 3.6 Flash (Reasoning, Minimal)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
6.3s
Cost:
$0.005403
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
9.61%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
4.8041%
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 this operational nightmare, Molly whispered, shoving a hand-drawn map across the sticky coffee shop table. Bill squinted, adjusting his fake moustache. Is that a bouncy castle or a crude drawing of a lung? It is the bouncy castle, Molly hissed. It represents our primary drop zone. We slide down the chute, snag the loot, and melt into the suburbs. Molly, you’ve lost your mind, Bill said, sipping his lukewarm espresso. They’re five-year-olds. We steal from hedge funds, not toddlers. This isn't just any toddler, Molly insisted, tapping the center of the paper. This is Timmy Montgomery. His father is a tech billionaire who thinks love is measured in luxury consumer goods. That pile of presents is worth at least forty grand on the black market. Forty grand of what? Bill asked. Play-Doh? Pokeball cards? Try a custom-built, diamond-encrusted electric mini-Ferrari, Molly said. Plus three unreleased gaming consoles and a small, gold-plated drone. I hate that I’m listening to you, Bill sighed. What’s the security? Brutal, Molly admitted. Five moms on high alert, three sugar-crazed sugar-fiends, and a mercenary named Barnaby. Who's Barnaby? The clown, Molly said grimly. I checked his credentials. He’s retired SWAT. He balloon-twists with military precision. Great, an tactical entertainer, Bill muttered. How do we get past the front gate? Disguises, Molly replied, pulling two oversized cardboard boxes from her bag. Absolutely not, Bill said. I am a master of stealth, not a box. You’re a mascot, Molly corrected. You’re Barnaby’s assistant, Bongo the Sad Hippo. I’m the catering staff bringing in the artisanal juice boxes. I hate Bongo, Bill snapped. Why do I always get the humiliated roles? Because you have the posture of a depressed mammal, Molly said. Now, listen. At fourteen hundred hours, the sugar crash hits. Timmy will melt down over a dropped cupcake. The mothers will rush to cover up the public embarrassment. That is our window. And the clown? Bill asked. I’ll neutralize Barnaby with a glitter bomb, Molly said smoothly. You're going to glitter-bomb a veteran? It’s humane, Molly argued. It blinds him with sparkle for precisely sixty seconds. You grab the mini-Ferrari, load the smaller boxes into the leather seats, and drive it straight through the wooden fence. Drive the kid's toy car through a fence? Bill asked, his eyes widening. It has forty horsepower, Bill. It's basically an Italian sports car. I’ll meet you at the getaway van with the juice box money. What if the kids swarm us? Bill asked, genuinely terrified. They’re relentless, Molly. They smell fear. And frosting. If they approach, throw the juice boxes, Molly commanded. It creates a feeding frenzy. They’ll turn on each other. You are a monster, Molly, Bill said, shaking his head in reluctant admiration. A absolute criminal genius. I know, Molly smiled. Now put on the hippo head. We have a birthday party to ruin.
created_at
"2026-07-25T14:28:55.383Z"
7/25/2026, 2:28:55 PM
model
"google/gemini-3.6-flash-20260721"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
627
moderation_latency
(null)
generation_time
6212
tokens_prompt
75
tokens_completion
762
native_tokens_prompt
67
native_tokens_completion
707
native_tokens_completion_images
0
native_tokens_reasoning
0
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
0.005403
router
(null)
provider_responses
0
endpoint_id
"5d6d133d-a953-4781-a835-7ee76f4e1388"
id
"J8hkapPZG9uc1MkPm_-XqQI"
is_byok
false
latency
627
model_permaslug
"google/gemini-3.6-flash-20260721"
provider_name
"Google AI Studio"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784989735-kVjn4eHFJwDjiur36pks"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784989735-NJkqR2MFHPKFAQ0zVoo6"
upstream_id
"J8hkapPZG9uc1MkPm_-XqQI"
total_cost
0.005403
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
9.61%
Matches word count
n/a
neededClean
false
words
478
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
491
wordsDialogue
0
4.8041%