NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-10-01KR8X2HCK36W4CQ48GPW87EJB
Run Details
Model:
Gemini 3.1 Flash Lite
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
3.8s
Cost:
$0.000906
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
34.0511%
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 minivan idled at the curb while Molly adjusted her balaclava, which was currently pinching her nose in a way that made her look like a grumpy pug. Beside her, Bill was frantically smoothing down his tuxedo, an outfit he had insisted was required for a stealthy intrusion into a gated suburban birthday bash. The target was the Miller residence, specifically a pristine stack of high-end loot resting on a mahogany table in the living room. Bill, are you sure about the clown shoes, Molly asked, squinting at his oversized footwear. They are standard infiltration gear for this environment, he replied, checking his watch with exaggerated gravity. If we get spotted, we blend in with the entertainment until the coast is clear. It is brilliant. Molly sighed, leaning back against the passenger door. We are robbing a seven-year-old, Bill. Do we really need a contingency plan involving slapstick comedy? I just want the pile of gift cards and the fancy electronics. Bill shook his head, his bow tie bobbing. Precision is the soul of a good heist. We go in through the garden gate, bypass the bounce house, and hit the gift table during the cake cutting ceremony. It is foolproof. The engine hummed as silence stretched between them. Through the gaps in the tall oak trees, they could hear the muffled squeals of children fueled by too much sugar. Molly checked her watch. Ten minutes until the candles were lit. The logistics were simple, yet Bill insisted on dressing like a disgruntled circus performer. She grabbed her tactical bag, which was stuffed with empty pillowcases and a very sophisticated lock-picking kit that lived permanently next to a bag of beef jerky. Is the cake chocolate or vanilla, Bill asked suddenly. Who cares, Molly snapped. Focus on the layout. Where is the dog? The dog is a golden retriever named Buster, Bill noted, consulting a crumpled map. He is currently preoccupied with a rogue juice box under the patio furniture. We have a clear path to the French doors. Good, Molly said, grabbing the handle. If we get caught, I am telling them you are the hired help and I am just the disgruntled aunt who lost a bet. Bill adjusted his frilly cuffs. I prefer the term master of disguise, but I suppose that works. Just try not to trip over your own heavy breathing when we run toward the getaway car. They stepped out into the manicured grass, the suburban sunset casting long, orange shadows across their path. The house looked peaceful, a fortress of wrapping paper and bows awaiting their unwanted arrival. Molly moved with practiced fluidity, staying low, while Bill clomped along behind her in his giant shoes, looking like a depressed penguin. This was it, the high-stakes world of nursery-level grand larceny, and they were ready to snatch the spoils before the final chorus of happy birthday echoed into the night.
created_at
"2026-05-10T12:16:28.059Z"
5/10/2026, 12:16:28 PM
model
"google/gemini-3.1-flash-lite-20260507"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
595
moderation_latency
(null)
generation_time
3721
tokens_prompt
77
tokens_completion
729
native_tokens_prompt
67
native_tokens_completion
593
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
"standard"
usage
0.00090625
router
(null)
provider_responses
0
endpoint_id
"237af107-d2e9-46c7-b146-c74d75980dee"
id
"HHcAavS4D4Wtz7IP88SG2QM"
is_byok
false
latency
595
model_permaslug
"google/gemini-3.1-flash-lite-20260507"
provider_name
"Google AI Studio"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1778415388-kLiKmD0QsSHGk4lOEk0a"
session_id
(null)
api_type
"completions"
id
"gen-1778415388-K2JGJ8F0cnWH9vXFygkh"
upstream_id
"HHcAavS4D4Wtz7IP88SG2QM"
total_cost
0.00090625
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
68.10%
Matches word count
n/a
neededClean
false
words
486
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 30.00%
neededClean
false
wordsTotal
492
wordsDialogue
0
34.0511%