NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY8XDHT20MCE40SR02RM7ZK
Run Details
Model:
Gemma 3 27B
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
10.9s
Cost:
$0.000065
Provider:
Novita
Evaluation summary
Score
Evaluator
Details Link
2.01%
Matches word count
(details)
98.11%
Dialogue to Total Word Ratio
(details)
50.0632%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
“Right, so glitter. Everywhere, apparently.” Molly said, picking a rogue speck off her leather jacket. “Glitter is the least of our worries. It’s the *ball pit*.” Bill groaned, sketching furiously in a notepad. “It’s a logistical nightmare. A swirling vortex of small, sticky children.” “We agreed, Bill, in and out. Grab the gifts, avoid direct contact. Like ninjas. Glitter-covered ninjas.” “Ninjas don’t have to navigate a sea of miniature humans attempting to drown each other in plastic balls. What’s the birthday kid into? Dinosaurs? Princesses? Knowing this informs our distraction strategy.” “Apparently, a very intense obsession with garden gnomes. His mother mentioned it on Facebook. We’re talking gnome-themed everything.” “Gnomes. Perfect. I’ve been practicing my gnome impression. It’s… surprisingly effective.” Bill puffed out his cheeks and mumbled, “Greetings, small humans, I am Reginald, keeper of the garden…” “Please, no. Just… no. Focus on the presents. How many are we talking?” “Estimates put it at around thirty. Mostly Lego, a remote-controlled car, and a concerning amount of Play-Doh.” “Thirty presents. Worth it.”
created_at
"2026-02-20T19:37:33.51Z"
2/20/2026, 7:37:33 PM
model
"google/gemma-3-27b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
753
moderation_latency
(null)
generation_time
10859
tokens_prompt
93
tokens_completion
329
native_tokens_prompt
75
native_tokens_completion
280
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
0.000064925
router
(null)
provider_responses
0
endpoint_id
"cf530f36-7cb8-4eb9-8953-880eef952af3"
id
"114b87c9-4c4e-43f2-886c-9384ecc9bac0"
is_byok
false
latency
626
model_permaslug
"google/gemma-3-27b-it"
provider_name
"Novita"
status
200
api_type
"completions"
id
"gen-1771616253-t6EvVM4JqowZIgSWzbXQ"
upstream_id
"114b87c9-4c4e-43f2-886c-9384ecc9bac0"
total_cost
0.000064925
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Novita"
Evaluation details
Result
Evaluator
Details
Meta Data
2.01%
Matches word count
n/a
neededClean
false
words
175
98.11%
Dialogue to Total Word Ratio
Ratio: 86.29%, Deviation: 3.71%
neededClean
false
wordsTotal
175
wordsDialogue
151
50.0632%