NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-03-01KJV0PM1VK0J65AQ4R39ZZKP3
Run Details
Model:
Gemini 3.1 Flash Lite (Preview)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
3.5s
Cost:
$0.000947
Provider:
Google AI Studio
Evaluation summary
Score
Evaluator
Details Link
97.63%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
48.8138%
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 adjusted her neon pink tutu and squinted at the blueprints spread across the dashboard of the rusted van. The floor was littered with glitter and discarded juice boxes. Bill sat in the driver seat, meticulously attaching a fake mustache to his upper lip with chewing gum. He looked less like a master criminal and more like a confused walrus. Are you sure the inflatable obstacle course is the weak point, Bill? It is the only way to bypass the laser grid of tethered balloons, Molly. But the itinerary says the birthday boy is six. Can a six-year-old really afford a high-end security system? You would be surprised by the budget for these suburban princess-themed bashes. They looked out the windshield at the backyard sanctuary. Three dozen kids were currently shrieking in a chaotic frenzy of sugar-induced delirium. A pile of glitter-wrapped loot sat unguarded near a folding table. The perimeter guard is a golden retriever wearing a tiara. Is that going to be an issue? Dogs are simple creatures, Bill. Just toss him a stray cupcake and he will abandon his post immediately. Fair enough. What about the parents? They are currently engaged in a heated debate about organic kale chips. They will not notice us if we stay low and pretend to be part of the entertainment staff. I hate wearing these oversized clown shoes. They make it impossible to execute a silent tactical roll. Stop complaining. You look festive. Are you ready to execute Operation Sugarplum? I am ready. Do we have the reinforced pillowcases for the haul? Yes. I have them hidden under this ridiculous cape. Make sure you grab the boxes that rattle the loudest. Those are the ones with the high-end electronics. Understood. What is our exit strategy if the clown shows up for his shift? We hit him with the bag of confetti and flee through the hedge maze. That seems unnecessarily dramatic, but I like the theatrics. Let us do this, Molly. Wait, check your mustache one more time. It is peeling off near the corner of your mouth. Dash it all. Hand me the backup gum. Molly tossed him a piece of bubblicious. She checked her watch and took a deep breath. The target was ripe for the picking, and the children were far too distracted by a magical show involving bubbles to notice two adults lurking in the shadows of the bounce house. Grab your cape, Bill. It is time to liberate those toys from the clutches of childhood. Roger that. If we are back in the van within three minutes, we can make it to the taco stand before they close. Ambition is the mark of a true professional. They climbed out of the van. Bill tripped over his clown shoes, stumbled into a hydrangea bush, and accidentally initiated a somersault. Molly rolled her eyes and marched toward the glittery pile, her tulle skirt rustling with the quiet promise of grand larceny.
created_at
"2026-03-03T23:32:00.712Z"
3/3/2026, 11:32:00 PM
model
"google/gemini-3.1-flash-lite-preview-20260303"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
516
moderation_latency
(null)
generation_time
3357
tokens_prompt
77
tokens_completion
747
native_tokens_prompt
67
native_tokens_completion
620
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)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"STOP"
usage
0.00094675
router
(null)
provider_responses
0
endpoint_id
"e004ce39-eb0c-4474-93d5-55072e725970"
is_byok
false
latency
516
model_permaslug
"google/gemini-3.1-flash-lite-preview-20260303"
provider_name
"Google AI Studio"
status
200
api_type
"completions"
id
"gen-1772580720-7Q8J9224s8Qy3W0neZeX"
upstream_id
(null)
total_cost
0.00094675
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Google AI Studio"
Evaluation details
Result
Evaluator
Details
Meta Data
97.63%
Matches word count
n/a
neededClean
false
words
493
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 50.00%
neededClean
false
wordsTotal
500
wordsDialogue
0
48.8138%