NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCYM3W74T4C9C48KRN78MGS
Run Details
Model:
Gemini 3.6 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
1m 8s
Cost:
$0.126423
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.80%
Dialogue to Total Word Ratio
(details)
99.9001%
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 unrolled the master blueprint across the grease stained kitchen table. It was a crude crayon drawing of a suburban backyard, complete with a smiling yellow sun in the top corner. Bill leaned in, his heavy night vision goggles resting atop his forehead, squinting through a dense cloud of burnt toast smoke at the elaborate tactical diagram. "This is the drop zone, Bill. Little Timmy turns seven at two o'clock sharp. By three o'clock, the main loot table will be completely unguarded, stacked six feet high with Lego sets, remote controlled helicopters, and at least three envelopes containing unspeakable cash from his wealthy grandmother." Bill pulled a magnifying glass from his trench coat and carefully examined the sketch. He pointed a trembling finger at a bulbous, hand drawn shape painted right near the sandbox. "What about the inflatable fortress of doom? You know I do not do heights, Molly. Or vinyl slide friction burns. Remember the bouncy castle disaster in Des Moines?" Molly snapped loudly, pacing the linoleum kitchen floor with intense focus and frustration. She slammed a cardboard box of apple juice directly onto the center of the map. "That was a rogue bounce house, and you know it. We drop in under cover of the magic show. While Barnaby the Clown distracts the miniature heathens with balloon poodles, you bypass the juice bar, grab the glittery packages, and head for the fence. Simple." Bill adjusted his dark tactical turtleneck and shook his head in absolute disbelief. He pulled out a stopwatch, clicking it with dramatic flourish before unholstering a shiny silver party popper. "It is never simple, Molly! What about the mother? Mrs. Gable is a ruthless operative. I saw her refilling the snack bowls earlier. She wears juice pouches on her belt like live grenades! If she catches us without a party hat, we are dead meat." Molly pulled two sparkly, cone shaped party hats from her heavy duffel bag and aggressively tossed one directly at his broad chest, knocking his magnifying glass onto the counter. "Which is why we blend in. You put on the glitter cone, grab a handful of goldfish crackers, and act like a very tall, heavily bearded child. Once the pinata explodes, it is total chaos. We sweep the gift table and disappear." Bill carefully tied the elastic chin strap of the pink party hat under his bushy beard. He looked into the cracked wall mirror, adjusting the fluffy pom pom on top with extreme focus and intense care. "Fine, but I get the giant Xbox package. And if there is a chaotic pinata swarm, I am definitely using you as a human shield, Molly. Deal? Now let us go out there right now and steal some prime childhood memories today." Molly grinned, strapping on her black ninja mask directly over her pointy party hat. She checked her digital stopwatch, adjusted her grip on an empty velvet loot bag, quietly unlocked the backdoor latch, and nodded toward the sunny, grassy suburban backyard.
created_at
"2026-07-25T15:34:41.295Z"
7/25/2026, 3:34:41 PM
model
"google/gemini-3.6-flash-20260721"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1394
moderation_latency
(null)
generation_time
67500
tokens_prompt
75
tokens_completion
4590
native_tokens_prompt
67
native_tokens_completion
16843
native_tokens_completion_images
0
native_tokens_reasoning
16224
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
0.126423
router
(null)
provider_responses
0
endpoint_id
"5d6d133d-a953-4781-a835-7ee76f4e1388"
id
"kddkatrTFcSs-8YPuaniwAM"
is_byok
false
latency
1394
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-1784993681-jSCTIMWfHEPNPcmbt5X0"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784993681-tLwwh2hkNWNXyvP5i6sI"
upstream_id
"kddkatrTFcSs-8YPuaniwAM"
total_cost
0.126423
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
100.00%
Matches word count
n/a
neededClean
false
words
500
99.80%
Dialogue to Total Word Ratio
Ratio: 49.80%, Deviation: 0.20%
neededClean
false
wordsTotal
500
wordsDialogue
249
99.9001%