NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHYB0S48GHPC3QJNMTT23JPJ
Run Details
Model:
Gemma 3 12B
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
5.7s
Cost:
$0.000037
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
20.19%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
10.0954%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 community hall smelled overwhelmingly of sugar and desperation – desperation, in Molly's case, for a decent payday. Streamers drooped like forgotten dreams from the ceiling, a pathetic confetti graveyard. Bill, ever the pragmatist, was meticulously sketching the room’s layout on a napkin pilfered from a nearby refreshment table. Tiny pastel chairs sat arranged in neat rows, facing a stage adorned with a frankly alarming number of balloons. He circled a particularly large pile of brightly wrapped boxes with a thick red pen. “That’s the jackpot, I reckon,” he mumbled, more to the napkin than to her. Molly kicked at a rogue bouncy castle, the air whooshing out with a mournful sigh. “Kids,” she muttered. Just miniature, sugar-fueled chaos agents. Bill tapped the napkin. “Security's…minimal. One volunteer manning the door, looks like he moonlighted as a scarecrow.” “Right, because terrorizing five-year-olds is a rigorous training ground for security professionals,” Molly deadpanned. “So, inflatable diversion?” Bill nodded, pointing at the partially deflated bouncy castle. “Distraction, then swift retrieval. Present count is…significant.” “Let's just hope none of them have bite.”
created_at
"2026-02-20T20:14:20.82Z"
2/20/2026, 8:14:20 PM
model
"google/gemma-3-12b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
379
moderation_latency
(null)
generation_time
5521
tokens_prompt
93
tokens_completion
321
native_tokens_prompt
75
native_tokens_completion
264
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
0.00003732
router
(null)
provider_responses
0
endpoint_id
"5aef2a85-3d1c-4659-a522-98bbb5272c6a"
is_byok
false
latency
35
model_permaslug
"google/gemma-3-12b-it"
provider_name
"Crusoe"
status
429
1
endpoint_id
"eb06dc92-5a16-47ec-a776-6ef956457c47"
id
"chatcmpl-RDne3483kfZDYWgThcsIdxJX"
is_byok
false
latency
50
model_permaslug
"google/gemma-3-12b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771618460-aE0GYsmEEFRGyPVk97if"
upstream_id
"chatcmpl-RDne3483kfZDYWgThcsIdxJX"
total_cost
0.00003732
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
20.19%
Matches word count
n/a
neededClean
false
words
180
0.00%
Dialogue to Total Word Ratio
Ratio: 28.33%, Deviation: 18.33%
neededClean
false
wordsTotal
180
wordsDialogue
51
10.0954%