NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-11-19-01M7SPDXKF0RQYMQTXZ7RG2T15
Run Details
Model:
Gemini 4 Neon (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
55.5s
Cost:
$0.090276
Provider:
Google
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.95%
Dialogue to Total Word Ratio
(details)
99.9748%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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 tapped a red dry-erase marker against a sprawling blueprint taped to the hood of a rusted sedan. The diagram was a masterpiece of criminal cartography, detailing every shrub, sprinkler head, and patio chair in the Henderson backyard. "We breach the perimeter at fourteen hundred hours," Molly whispered. "That is precisely when the sugar crash hits. The targets will be sluggish, distracted by cake, and completely vulnerable to a swift flank maneuver." Bill squinted at the map through oversized binoculars he had worn since their early morning reconnaissance. He slowly lowered them, revealing a face painted with deep existential dread and a smear of powdered sugar. "Molly, the targets are seven years old," Bill replied, waving a half-eaten cruller. "We used to rob jewelry stores. Now we are plotting a tactical strike on a bouncy castle? What happens if a feral toddler bites my ankle?" Molly sighed and slid a glossy surveillance photograph across the dented hood. It showed a mountain of vibrantly wrapped boxes stacked high on a folding table, gleaming like dragon gold. "Look at the guest list, Bill," Molly insisted. "Little Preston Henderson is the heir to a massive pharmaceutical fortune. Those are not cheap plastic dinosaurs in those boxes. We are looking at vintage gaming consoles, rare sneakers, and at least three expensive drones." Bill leaned closer, his fragile moral compass spinning wildly before snapping directly toward greed. He poked a blurry figure in the corner of the photograph, his brow furrowing with sudden professional concern. "Okay, the loot is solid, I admit it," Bill muttered. "But what about the heavy security? The intel says they hired a guy named Barnaby the Magnificent. Magicians have quick hands, Molly. He could easily spot a palmed gift bag from twenty paces." Molly grinned wickedly and hoisted a battered canvas duffel bag onto the car, unzipping it to reveal neon spandex, a giant red foam nose, and a portable helium tank. "Barnaby is going to be mysteriously delayed by a flat tire," Molly declared proudly. "Which means the party will need a frantic last-minute replacement. You are going in undercover as a balloon artist, and I am driving the getaway minivan." Bill stumbled backward, tripping over a rusty toolbox and nearly dropping his sugary pastry into a puddle of motor oil. He stared at the shimmering spandex as if it were a venomous snake. "I am not wearing purple tights in broad daylight," Bill shouted. "I have a serious reputation in the underworld! If the guys from the syndicate see me twisting latex into a poodle, I will never live it down." Molly calmly plucked a squeaky polka-dot bowtie from the open bag and tossed it squarely at his chest. She rolled up the blueprints with the crisp efficiency of a seasoned general preparing for war. "Relax, Bill," Molly chuckled. "You only have to make one decent giraffe, swipe the gaming console, and sprint for the hedge. Now put on the giant shoes before the bounce house deflates."
created_at
"2026-11-19T10:42:53.691Z"
11/19/2026, 10:42:53 AM
model
"google/gemini-4-neon-20261118"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1928
moderation_latency
(null)
generation_time
55402
tokens_prompt
75
tokens_completion
920
native_tokens_prompt
97
native_tokens_completion
5999
native_tokens_completion_images
0
native_tokens_reasoning
5375
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
0.090276
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.2; linux; x64))"
http_referer
(null)
request_id
"req-1795084973-r5xeDce02CG6IR8Ev3wK"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1795084973-69lEwfDeVR6O74LnuqBv"
upstream_id
"rdL-auOAMNqH9LsPt7fYwAg"
provider_responses
0
endpoint_id
"cbc1d01f-e431-41ec-9abb-36c1a5b10cb4"
id
"rdL-auOAMNqH9LsPt7fYwAg"
is_byok
false
latency
1928
model_permaslug
"google/gemini-4-neon-20261118"
provider_name
"Google"
status
200
total_cost
0.090276
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
499
99.95%
Dialogue to Total Word Ratio
Ratio: 49.90%, Deviation: 0.10%
neededClean
false
wordsTotal
503
wordsDialogue
251
99.9748%