NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-11-19-01M7SPT9EN1DN7SDV5VVV3MH3C
Run Details
Model:
Gemini 4 Neon (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
33.1s
Cost:
$0.054051
Provider:
Google
Evaluation summary
Score
Evaluator
Details Link
97.63%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
98.8138%
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
Molly slapped a glossy brochure from Bounce-A-Lot Emporium onto the scarred kitchen table. Above them, a bare bulb swung gently, casting dramatic shadows over a blueprint of a suburban backyard. Red yarn connected a blurry photograph of a bouncy castle to a sticky note reading The Perimeter. Bill sat across from her, chewing a stale pretzel stick while staring at the layout with deep, existential dread. He had cracked safes in Zurich and bypassed lasers in Monaco, yet here they were. "The target is Timmy Henderson," Molly said, tapping the brochure with a laser pointer. "Turning seven. Rich parents. We are looking at thirty high-end Lego sets and possibly a motorized scooter." Bill sighed and leaned back in his creaky wooden chair. Molly had been awake for forty-eight hours, fueled entirely by cheap espresso and the thrill of petty larceny. She wore a black turtleneck despite the sweltering summer heat, fully committed to the aesthetic of a high-stakes criminal mastermind. "Molly, this is a kid's birthday party," Bill replied, gesturing toward the diagram. "What happened to art galleries? What happened to diamond vaults? You want me to rappel down a maple tree for plastic bricks?" Molly glared at him, pacing the tiny linoleum floor. To her, a score was a score, regardless of whether the security detail consisted of armed guards or overworked mothers holding lukewarm wine. The resale market for mint-condition toy trains was booming, and she refused to let Bill's sudden conscience ruin their retirement fund. "The Millennium Falcon set retails for eight hundred dollars, Bill," Molly snapped. "And the security is laughable. One hired clown who is definitely hungover, and a distracted dad manning the grill. We slip in during the pinata distraction, bag the loot, and vanish before they cut the cake." Bill rubbed his temples, feeling a massive headache forming behind his eyes. He looked down at the intricate timeline Molly had drawn in green marker. At two-fifteen, the singing of the birthday song would create acoustic cover for their breach of the sliding glass door. While the children were blinded by sugar and chaos, Bill was supposed to wheel a stolen wheelbarrow full of brightly wrapped boxes right through the side gate. "Fine," Bill muttered, circling the patio table on the blueprint. "I am in. But I am not wearing a disguise, and if there is a sheet cake with buttercream frosting, I am taking a corner piece for myself." "Deal," Molly said, grinning wildly. "Just make sure you grab a piece with a frosting flower on it for me, and do not drop the Death Star." She reached under the table and pulled out two oversized party hats, sliding one across the scratched wood toward her partner. They had three hours until the bouncy castle was fully inflated, and she intended to execute the greatest suburban heist in history. Bill stared at the tiny elastic chin strap, sighed one last time, and began packing his lockpicks.
created_at
"2026-11-19T10:49:39.038Z"
11/19/2026, 10:49:39 AM
model
"google/gemini-4-neon-20261118"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1960
moderation_latency
(null)
generation_time
33113
tokens_prompt
75
tokens_completion
1522
native_tokens_prompt
97
native_tokens_completion
3584
native_tokens_completion_images
0
native_tokens_reasoning
2965
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
0.054051
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.22.2; linux; x64))"
http_referer
(null)
request_id
"req-1795085379-95glG2Zv91QHE4YfWMbU"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1795085379-Dg0csBKdx8NW38zg4Dz7"
upstream_id
"Q9T-aoG-B9Kjq8YPgb_ryAs"
provider_responses
0
endpoint_id
"cbc1d01f-e431-41ec-9abb-36c1a5b10cb4"
id
"Q9T-aoG-B9Kjq8YPgb_ryAs"
is_byok
false
latency
1960
model_permaslug
"google/gemini-4-neon-20261118"
provider_name
"Google"
status
200
total_cost
0.054051
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
97.63%
Matches word count
n/a
neededClean
false
words
493
100.00%
Dialogue to Total Word Ratio
Ratio: 30.00%, Deviation: 0.00%
neededClean
false
wordsTotal
500
wordsDialogue
150
98.8138%