NC Bench
Overview
About
Tests
Models
Vendors
Compare
Run: 2026-02-25-01KJ94VBEJJW4H5CMCTKK79JZF
Run Details
Model:
Gemini 2.5 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
1m 31s
Cost:
$0.048543
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
70.42%
Dialogue to Total Word Ratio
(details)
35.2097%
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 ran a skeptical finger over the crumpled blueprint spread across their rickety card table. Gloom filtered through a grimy window, illuminating dust motes dancing in the air, a stark contrast to the vibrant chaos depicted on the blueprint for a child's birthday. "Are you absolutely *sure* this is the score, Bill?" Bill, ever the connoisseur of questionable snacks, ignored her. He leaned back in his creaky chair, crunching loudly on a bright orange cheese puff, clouds of dust billowing from his opulent mustache. "Absolutely, Molly-kins. Prime target. High yield. Low risk." "Low risk?" Molly's voice was dry. "It's a six-year-old's birthday party. Security? Sparky the golden retriever and a perpetually sullen teenager." She tapped a crayon drawing of a crudely rendered bouncy castle, a sardonic brow raised. "Exactly!" Bill gestured with an orange-stained finger. "No laser grids. No pressure plates. Just... pure, unadulterated joy. And presents." Molly pinched her nose. "Objective: 'presents.' Not 'joy.' How many? Twenty? Half of them arts and crafts?" "Ah, but you forget. These are the Hastings' children. Multi-millionaire Hastings. Little Petunia's party is a NASDAQ opening bell for designer toys." Bill pushed the blueprint closer, his eyes gleaming. "The 'Gift Emporium.' A separate display room. Prime." Molly's eyes narrowed. "A 'Gift Emporium' for a six-year-old? Right. What's our approach? Stealth? Bribe the clown?" "Bribing the clown is an option," Bill mused. "But I was thinking 'Trojan Pony.' We infiltrate as entertainment." Molly stared, disbelief clear. "Entertainment? Bill, we're thieves. You can barely do a cartwheel." "Details! 'Magicians.' You, my assistant, handle the vanishing act. With the presents." "So, I distract the kids, you charm the parents, we vanish with two dozen Barbie dream houses?" "Precisely! Gifts into reinforced duffel bags. Disguise: stressed party planner. Nobody questions a stressed planner." "And you?" "I'll be 'Amazing Bartholomew.' Top hat, cape. Smoke bomb for my grand exit, obviously." Molly sighed. "You're setting off a smoke bomb at a children's party, aren't you?" "Only if necessary! Dramatic flair. Or if Sparky gets overly enthusiastic. Timeline? Nap-time?" "Petunia and cronies well-napped by three. Present opening at four. Narrow window, post-cake sugar rush, pre-tantrum meltdown." "Perfect. Chaos is our cover," Bill clapped his hands, scattering cheese puff dust. "Getaway vehicle? Ice cream truck again?" "Bill, no. Not the ice cream truck. Too much attention, tunes stick in my head." Molly jabbed at a marked spot on the map. "The discreet minivan. Tinted windows. Plenty of space. No jingle-jangle." "Spoilsport," Bill grumbled. "Minivan it is. But I'm still getting a smoke bomb. Non-toxic. We're professionals, Molly. Standards."
created_at
"2026-02-25T00:58:10.265Z"
2/25/2026, 12:58:10 AM
model
"google/gemini-2.5-flash"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1350
moderation_latency
(null)
generation_time
90952
tokens_prompt
77
tokens_completion
7958
native_tokens_prompt
67
native_tokens_completion
19409
native_tokens_completion_images
0
native_tokens_reasoning
18763
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"STOP"
usage
0.0485426
router
(null)
provider_responses
0
endpoint_id
"16c01e26-9b91-4d22-b487-ed35634b017d"
is_byok
false
latency
1350
model_permaslug
"google/gemini-2.5-flash"
provider_name
"Google AI Studio"
status
200
api_type
"completions"
id
"gen-1771981090-jN3S84Rq5gBI8iMwXP4k"
upstream_id
(null)
total_cost
0.0485426
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
426
70.42%
Dialogue to Total Word Ratio
Ratio: 67.35%, Deviation: 2.65%
neededClean
false
wordsTotal
438
wordsDialogue
295
35.2097%