NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCT9H3RKACQ8Q49F922XX7S
Run Details
Model:
Gemini 3.6 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
26.9s
Cost:
$0.049376
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
99.92%
Matches word count
(details)
99.77%
Dialogue to Total Word Ratio
(details)
99.8447%
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 blueprint onto the greasy kitchen table, sending two cold slices of pepperoni pizza sliding onto the floor. She did not care. This was the big one. She tapped the paper with a red marker, her eyes wide with the manic enthusiasm of a tactical criminal mastermind who had entirely lost her sense of scale. She adjusted her night-vision goggles, which were entirely unnecessary in the broad daylight of two o'clock in the afternoon. Bill leaned back in his squeaky folding chair, chewing nervously on a sour gummy worm. He glanced at the scribbled notes. Red arrows pointed toward a bouncy castle, and a heavily annotated diagram depicted a mercenary labeled as Barnaby the Clown. "Look at the perimeter, Bill. It is an absolute fortress," Molly whispered, pointing at a crayon drawing of a suburban split-level house. Bill adjusted his fake mustache, which was already losing its adhesive in the humid apartment air. He rubbed his temples in total exhaustion. "Molly, it is a seven-year-old boy's birthday party," Bill said. "The security detail consists of a guy in squeaky shoes and a mother who forgot the juice boxes." Molly slammed her fist on the table, her tactical headset dropping around her chin. She adjusted it with a fierce scowl. She checked the tension on her miniature grappling hook, which was originally intended for a heavy-duty dog leash. Never underestimating the enemy was rule number one in the Heist Codex, a spiral notebook she had bound herself using stolen duct tape. "Barnaby is a wildcard, Bill! He has balloon animals, which means he carries concealed rubber. And Mrs. Higgins has three dogs. Teacup poodles, Bill. Loud, aggressive, highly alert." Bill sighed, dropping the half-eaten gummy worm onto a stack of glittery party hats. He looked out the window at the sunny suburban cul-de-sac. The target house across the street was decorated with pastel streamers and a giant banner that read Happy Birthday Little Timmy. "We are grown adults planning to steal Lego sets and a motorized ride-on jeep," Bill pointed out, crossing his arms tightly. "Is this really the grand crown jewel of our criminal career?" Molly grabbed him by his denim jacket, pulling his face inches from hers. Her eyes burned with a dark, unhinged ambition. She had spent three weeks scouting the local toy store, tracking the online registry, and calculating the exact cubic volume of the gift table located in the sunroom. "That motorized jeep has custom leatherette seats and four-wheel drive, William! We extract the entire loot pile during the blindfolded pinata phase. Total chaos. Zero eyes on the prize." Bill blinked, considering the sheer tactical absurdity of the proposition. He checked his plastic wristwatch. The party started in fifteen minutes. He grabbed a smoke grenade shaped like a juice box. "Fine," Bill muttered. "But if I get bitten by a poodle or tackled by a clown, I am telling the cops you forced me into this." "Fair enough," Molly replied. "Now put on your camouflage party hat."
created_at
"2026-07-25T14:19:00.095Z"
7/25/2026, 2:19:00 PM
model
"google/gemini-3.6-flash-20260721"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1336
moderation_latency
(null)
generation_time
26874
tokens_prompt
75
tokens_completion
2903
native_tokens_prompt
67
native_tokens_completion
6570
native_tokens_completion_images
0
native_tokens_reasoning
5922
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
0.0493755
router
(null)
provider_responses
0
endpoint_id
"5d6d133d-a953-4781-a835-7ee76f4e1388"
id
"1MVkaru4DPLTjMcP5Zr64As"
is_byok
false
latency
1336
model_permaslug
"google/gemini-3.6-flash-20260721"
provider_name
"Google AI Studio"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784989140-w7XHSKkZyyiqkxfpaqT2"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784989140-0B9Ny1X0KaAPmj3ojmiF"
upstream_id
"1MVkaru4DPLTjMcP5Zr64As"
total_cost
0.0493755
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
99.92%
Matches word count
n/a
neededClean
false
words
503
99.77%
Dialogue to Total Word Ratio
Ratio: 30.21%, Deviation: 0.21%
neededClean
false
wordsTotal
513
wordsDialogue
155
99.8447%