NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY7Z80ST4VZZR45Q5QKBBQV
Run Details
Model:
Gemma 3 12B
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
8.5s
Cost:
$0.000036
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
98.43%
Dialogue to Total Word Ratio
(details)
83.2647%
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 bouncy castle loomed, a garish pink behemoth pulsating with shrieks and the frantic thumping of tiny feet. Molly chewed nervously on a strand of her fiery red hair, squinting at the scene. Bill, meanwhile, was meticulously sketching a diagram in the dirt with a twig, occasionally pausing to meticulously rearrange a nearby pebble. He’d decided the inflatable slide was the key, a perfect distraction. A small girl in a tiara ran past, narrowly avoiding a collision with Bill's leg. He didn’t even flinch, maintaining his focused gaze upon the diagram. Streamers, balloons, an astonishing quantity of brightly wrapped boxes – it was a veritable mountain of loot. He estimated at least sixty presents, possibly more. The sheer volume was almost overwhelming. Molly kicked at a stray party favour, a miniature plastic pirate ship. “Think they’ll miss them?” she mumbled. Bill offered a brief, dismissive wave without looking up. He added a tiny, detailed illustration of a clown. "Precision, Molly. It's all about precision." He circled the clown three times with his twig. "Minimum disruption." He then meticulously erased the clown with the side of his thumb.
created_at
"2026-02-20T19:21:04.802Z"
2/20/2026, 7:21:04 PM
model
"google/gemma-3-12b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
337
moderation_latency
(null)
generation_time
8434
tokens_prompt
93
tokens_completion
305
native_tokens_prompt
75
native_tokens_completion
251
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
0.00003563
router
(null)
provider_responses
0
endpoint_id
"eb06dc92-5a16-47ec-a776-6ef956457c47"
id
"chatcmpl-RLn1WsU1Jm1zI55QDufJYB3m"
is_byok
false
latency
58
model_permaslug
"google/gemma-3-12b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771615264-euTsFl1nMvkRxFTG0hRF"
upstream_id
"chatcmpl-RLn1WsU1Jm1zI55QDufJYB3m"
total_cost
0.00003563
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
68.10%
Matches word count
n/a
neededClean
false
words
186
98.43%
Dialogue to Total Word Ratio
Ratio: 6.45%, Deviation: 3.55%
neededClean
false
wordsTotal
186
wordsDialogue
12
83.2647%