NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-11-19-01M7SPXBMW452E8HXWY6F9J9ZY
Run Details
Model:
Gemini 4 Neon (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 9s
Cost:
$0.116451
Provider:
Google
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.81%
Dialogue to Total Word Ratio
(details)
99.9032%
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 slapped a laminated map of the Henderson backyard onto the card table. The bare bulb overhead swung slightly, casting dramatic shadows across the hand-drawn schematic of a bouncy castle and a kiddie pool. She tapped a red circle near the patio with a plastic ruler. Bill leaned over the table, squinting through a cloud of cheddar dust rising from his cheese puffs. "The target is the gift table," Molly said. "It sits right here, flanked by the snack station and the face-painting booth. High traffic, zero cover. We have a four-minute window while the marks watch the magician." "Hold on," Bill said. "What kind of magician are we talking about? Are we dealing with a guy pulling coins from ears, or a high-level illusionist who might spot us using mirrors?" Molly sighed and rubbed the bridge of her nose. She had pulled off museum heists in Prague, yet here she was, explaining basic recon to a man with orange fingers. She flipped through a manila folder and slid a blurry photograph across the table. It showed a middle-aged man in a rumpled tuxedo holding a tired rabbit. "His name is The Great Lorenzo," Molly said. "My bakery source says his card tricks are terrible, and he takes a smoke break precisely at three. That is when we strike. You disable the perimeter, and I bag the loot." "Wait a minute. How do I disable a seven-year-old's birthday party? Are there lasers? Do I need the glass cutter?" Bill reached into his duffel bag and pulled out heavy bolt cutters, looking hopeful. Molly snatched them away and shoved a bright red foam nose into his chest. Bill stared at the squeaky sphere in his palm as if it were a live grenade. "You are going in undercover as a backup clown," Molly said. "Create a diversion near the pinata. Juggle, fall over, release the helium balloons. Once the parents rush in to stop the crying, I sweep the table for the vintage Lego sets and that new gaming console." "I am a professional safecracker, Molly. I do not juggle. Also, kids are terrifying. They have sticky hands and zero respect for personal space. What if one of them kicks me in the shins?" Molly stepped closer, adjusting the collar of her black turtleneck. Her eyes gleamed with the cold, calculating fire of a true criminal mastermind. She picked up a roll of silver duct tape and tossed it into the duffel bag alongside a giant canvas sack. "If a kid kicks you, you take the hit," Molly said. "Do you know what an unopened Millennium Falcon goes for online? This score sets us up for the summer. Now put on the big shoes and practice your balloon animals." Bill groaned, slowly pulling a pair of oversized yellow polka-dot shoes from beneath the table. He slipped one onto his foot, the rubber squeaking loudly in the quiet, damp basement. "Fine," Bill muttered. "But I am keeping all the leftover birthday cake."
created_at
"2026-11-19T10:51:19.588Z"
11/19/2026, 10:51:19 AM
model
"google/gemini-4-neon-20261118"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2025
moderation_latency
(null)
generation_time
68805
tokens_prompt
75
tokens_completion
2215
native_tokens_prompt
97
native_tokens_completion
7744
native_tokens_completion_images
0
native_tokens_reasoning
7110
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
0.116451
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.2; linux; x64))"
http_referer
(null)
request_id
"req-1795085479-H9r0l6ND7G51DzNXg0wC"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1795085479-TfnpsnKAnP4gGUQ0peLY"
upstream_id
"p9T-asjkJ8enq8YPspib2QI"
provider_responses
0
endpoint_id
"cbc1d01f-e431-41ec-9abb-36c1a5b10cb4"
id
"p9T-asjkJ8enq8YPspib2QI"
is_byok
false
latency
2025
model_permaslug
"google/gemini-4-neon-20261118"
provider_name
"Google"
status
200
total_cost
0.116451
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
500
99.81%
Dialogue to Total Word Ratio
Ratio: 50.20%, Deviation: 0.20%
neededClean
false
wordsTotal
508
wordsDialogue
255
99.9032%