NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-05-20-01KS25WDT7S8RCYPX28M2XTEVT
Run Details
Model:
Gemini 3.5 Flash (Reasoning, Minimal)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
5.4s
Cost:
$0.006535
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
0.08%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0424%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
We have exactly four minutes once the bounce house goes up, Molly said, tapping her finger against a crudely drawn map of the backyard. That is the diversion. Bill squinted at the paper, adjusting his black ski mask which was currently bunching up around his ears. Why is the bounce house the diversion? Kids love bounce houses. They will be trapped inside it. Exactly, Molly said, rolling her eyes. They will be trapped in a state of gelatinous euphoria. The parents will be distracted by the safety hazard of forty sugar-loaded seven-year-olds leaping in unison. That is when we strike the gift table. I still think the clown disguise was a better play, Bill muttered. I bought the balloon animals, Molly. I practiced. I can make a sword. A slightly deflated sword, but a sword nonetheless. No clowns, Bill, Molly snapped. You have a fear of face paint and we both know you would end up crying in the ball pit. Now, focus. The target is the pile of wrapped boxes next to the juice box fountain. Is it a fountain of actual juice? Bill asked, his eyes widening. Like, cascading Capri Sun? It is irrelevant, Molly said. What is relevant is the haul. Little Timmy’s grandmother is loaded. We are talking high-end Lego sets, pristine gaming consoles, and possibly a motorized miniature sports car. We need a clean sweep. How do we carry a motorized sports car over a six-foot cedar fence? Bill asked. We drive it, Bill. We drive it through the flower beds. But we do not have the keys, he pointed out. It is a toddler’s toy, Bill! It operates on a giant red button. You do not need to hotwire a Barbie Dream Camper. Right, right, Bill said, nodding solemnly. Good planning. What about the guard? What guard? Molly frowned, looking at her map. There is no guard. The golden retriever, Bill said, pointing out the window toward the yard. He looks vigilant. He is currently chewing on a plastic lightsaber, but his eyes are cold, Molly. Cold and calculating. I brought a decoy steak, Molly sighed, pulling a slightly crushed ribeye from her pocket. Is that seasoned? Bill asked. Of course it is seasoned, I am not a monster, Molly said. Now, once the bounce house inflated blower starts, we scale the fence. You take the north side of the table, I take the south. We bag the loot in the heavy-duty sacks. What if there is a pinata? Bill asked suddenly. What about it? If there is a pinata, we should wait for the candy drop, Bill said earnestly. Free Snickers, Molly. It is a tactical acquisition. We are professional thieves, Bill, not trick-or-treaters, Molly hissed. We do not delay a high-stakes heist for fun-sized chocolate. It is peanut butter M&Ms, Molly. Think about the overhead. No, Molly said. We stick to the plan. Guard dog gets the steak, kids get the bounce house, we get the Lego. Are we clear? Bill sighed, adjusting his mask again. Clear. But if I see a red velvet cupcake, I am taking it. Fine, Molly said, checking her watch. Sync watches. Three, two, one... go.
created_at
"2026-05-20T07:51:25.773Z"
5/20/2026, 7:51:25 AM
model
"google/gemini-3.5-flash-20260519"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
813
moderation_latency
(null)
generation_time
5265
tokens_prompt
77
tokens_completion
788
native_tokens_prompt
67
native_tokens_completion
715
native_tokens_completion_images
0
native_tokens_reasoning
0
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
"standard"
usage
0.0065355
router
(null)
provider_responses
0
endpoint_id
"bd50061c-e1d8-469e-96b1-d322e77378f8"
id
"_WcNaqaMOe-h1MkPoY7bgAw"
is_byok
false
latency
813
model_permaslug
"google/gemini-3.5-flash-20260519"
provider_name
"Google AI Studio"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.15.0; linux; x64))"
http_referer
(null)
request_id
"req-1779263485-F24MW659qkKRk7nYq5On"
session_id
(null)
api_type
"completions"
id
"gen-1779263485-IQKc2BjVC6S3BgdkT9jN"
upstream_id
"_WcNaqaMOe-h1MkPoY7bgAw"
total_cost
0.0065355
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.08%
Matches word count
n/a
neededClean
false
words
529
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 70.00%
neededClean
false
wordsTotal
539
wordsDialogue
0
0.0424%