NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KRJT3NWH0PQBEAC0SMSGA8
Run Details
Model:
Gemini 3.7 Flash (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
35.3s
Cost:
$0.028835
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
43.38%
Matches word count
(details)
49.74%
Dialogue to Total Word Ratio
(details)
46.5582%
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 slammed the blueprint onto the metal hood of their battered surveillance van. The master layout had been hastily sketched in purple wax crayon across the back of a greasy diner placemat, complete with tiny detailed drawings of potted ferns, a sandbox, and an inflatable castle. "Walk me through the perimeter again, Bill," Molly demanded, firmly tapping a crooked square labeled Kitchen. "And do not leave out the hostile forces." Bill adjusted a pair of bulky night-vision goggles, which were entirely unnecessary inside the sweltering, sunlit vehicle parked along the quiet suburban cul-de-sac. He picked up an unsharpened pencil to use as an official tactical pointer. "At zero-eight-hundred hours, the target, code named Little Timmy, consumes three slices of double-fudge cake," Bill whispered, leaning over the steering wheel. "By zero-eight-fifteen, the sugar peaks. It will be utter anarchy in the living room." "Which gives us our primary extraction window," Molly muttered, nervously pacing the narrow space between two stacked metal toolboxes. "How many hostiles are we expecting on the ground?" "Twelve second-graders," Bill replied solemnly, consulting a heavily marked index card. "Three of them are biters, Molly. Completely untrained, hyperactive wildcards." "We neutralize them with the juice boxes," she said, kicking a heavy cardboard crate under the passenger bench. "Standard distraction protocol. What about the heavy perimeter security?" "The golden retriever is locked behind the laundry room baby gate," Bill confirmed, tapping an orange scribble on the map. "However, the hired entertainment remains an unknown variable." "Blinky the clown?" Molly scoffed, checking the latch on her duffel bag. "He folds under the slightest pressure. Where is the primary payload?" "The gift table sits beneath the dining room chandelier," Bill stated, squinting through a pair of cheap plastic binoculars. "Intel confirms two monster trucks, a remote-controlled drone, and four boxed Lego sets." "Those are black market goldmines for quick resale," Molly whispered, pulling a dark knit mask over her face with practiced precision. "What is our exact exfiltration route?" "We drop through the pet flap, stuff the loot into the laundry hampers, and scale the back fence before the piƱata breaks," Bill explained with immense pride. "Why on earth are you wearing a full tuxedo with a glittery purple cape?" Molly asked, glaring at his ridiculous outfit. "Magician disguise," Bill insisted, pulling a foam dove from his satin sleeve. "It provides theatrical cover if parents spot us." "Magicians do not rob suburban seven-year-olds," Molly snapped, tossing the fake bird out the window. "Stick to stealth maneuvers." "Fine, but I am keeping the cape," Bill grumbled, crossing his arms defensively. "It brings dramatic flair to grand larceny." "If that satin gets snagged on the bounce house zipper, you are on your own," Molly warned, tapping her digital watch. "Synchronize. In three minutes, we move." "Understood, boss," Bill said, pulling a black ski mask over his sequined top hat. "Let us go ruin Timmy's special day."
created_at
"2026-09-03T13:50:58.372Z"
9/3/2026, 1:50:58 PM
model
"google/gemini-3.7-flash-20260813"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
5562
moderation_latency
(null)
generation_time
30417
tokens_prompt
75
tokens_completion
2981
native_tokens_prompt
67
native_tokens_completion
7676
native_tokens_completion_images
0
native_tokens_reasoning
6994
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
0.02883525
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788443458-F0YHrP6ZGnSGMkvceB3a"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788443458-lz4q0tSoA4loZZMZzPo4"
upstream_id
"QnuZavL8H6DSjMcP0IWa2A0"
provider_responses
0
endpoint_id
"dd072012-3274-417d-97e3-e290cba37e31"
id
"QnuZavL8H6DSjMcP0IWa2A0"
is_byok
false
latency
5562
model_permaslug
"google/gemini-3.7-flash-20260813"
provider_name
"Google AI Studio"
status
200
total_cost
0.02883525
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
43.38%
Matches word count
n/a
neededClean
false
words
483
49.74%
Dialogue to Total Word Ratio
Ratio: 53.74%, Deviation: 3.74%
neededClean
false
wordsTotal
495
wordsDialogue
266
46.5582%