NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY7KBXD0QGWA29BJXVFRABS
Run Details
Model:
Gemma 3 27B
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
15.6s
Cost:
$0.000294
Provider:
Phala
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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
The chipped Formica of the diner booth felt cold under Molly’s elbows. Rain lashed against the window, mirroring the tempest brewing inside her. Opposite her, Bill, a man whose dedication to beige was frankly unsettling, meticulously separated the sugar granules in his packet. “Okay, so let’s run it again,” Molly said, stabbing a fork into a congealing pile of hash browns. “Little Timmy’s seventh birthday. ‘Pirate Adventure’ theme. Parents are… let’s call them ‘enthusiastically oblivious’. That’s from surveillance, by the way. Not my judgment. Though, accurate.” Bill finished his sugar segregation and began arranging the tiny white grains into a small pyramid. “Enthusiastically oblivious. Right. Good shorthand. Means they won't notice a slight…shifting of assets.” “Slight? Bill, we’re emptying a room of presents. It’s not a slight shifting. It’s a full-scale present evacuation.” “Semantics, Molly, semantics. Presentation is key. We want to give the impression everything is perfectly normal, just…lighter.” Bill finally looked up, his expression deadpan. “Think airlift, not smash and grab.” “Airlift? You’re suggesting we get a helicopter?” Molly choked on her orange juice. “No! Metaphorical airlift! We glide in, collect the goods, glide out. Clean. Efficient. Like a well-oiled, present-acquiring machine.” Molly rubbed her temples. “Right. The plan. You distract Captain Enthusiasm - that’s the dad, apparently has a full costume - I grab the presents. We’re aiming for maximum present density. Reports indicate a substantial Lego collection, a remote-controlled submarine, a surprisingly expensive drum kit…and a miniature pony.” Bill’s pyramid collapsed. He didn’t seem bothered. "A pony? They’re bringing a *pony* to a house party?" “Apparently. It's a petting zoo situation. Adds a layer of…complication." Molly sighed. "The parents are doing a ‘treasure hunt’ to direct the kids. That’s our window. Peak distraction.” "I’m thinking I'll challenge Captain Enthusiasm to a ‘walk the plank’ contest. Using a pool noodle, naturally. I’ve been practicing my pirate swagger." Bill puffed out his chest and attempted a menacing growl that sounded remarkably like a kitten. Molly stared. "Please, no swaggering. Just keep him occupied. Ask him about peg legs. Talk about buried treasure. Anything but swaggering. And focus on small talk, Bill. Don’t ask him if he’s ever considered a life of piracy." “But it’s relevant!” “No, it’s terrifying. You will freak out a seven-year-old’s father. And a potential witness.” Molly took a deep breath. "I grab the presents, we meet at the designated extraction point – Mrs Higgins’ prize-winning rose bushes – and then rendezvous at the van. Got it?” “Rose bushes? Really? That seems…vulnerable.” “It’s covered by a blind spot in the street camera. Trust me. Just… Bill, no sudden movements. No talking to the pony. And please, for the love of all that is holy, no piracy.”
created_at
"2026-02-20T19:14:35.577Z"
2/20/2026, 7:14:35 PM
model
"google/gemma-3-27b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
303
moderation_latency
(null)
generation_time
15415
tokens_prompt
93
tokens_completion
817
native_tokens_prompt
75
native_tokens_completion
714
native_tokens_completion_images
(null)
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
0.00029385
router
(null)
provider_responses
0
endpoint_id
"1d5251f0-02fd-484b-bb6c-d346277c8a0d"
id
"cmpl-bdfc2f84ad9e4f4b893bd8fe64ff4ae5"
is_byok
false
latency
263
model_permaslug
"google/gemma-3-27b-it"
provider_name
"Phala"
status
200
api_type
"completions"
id
"gen-1771614875-YnZ2MeT5Uz0LpatlpJFG"
upstream_id
"cmpl-bdfc2f84ad9e4f4b893bd8fe64ff4ae5"
total_cost
0.00029385
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Phala"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
454
0.00%
Dialogue to Total Word Ratio
Ratio: 73.57%, Deviation: 43.57%
neededClean
false
wordsTotal
454
wordsDialogue
334
0.0000%