NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHYAH2Z9M0GCA5CRVSSCYSS8
Run Details
Model:
Gemma 3 27B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
21.5s
Cost:
$0.000115
Provider:
DeepInfra
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 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
The chipped Formica of the diner booth felt sticky under Molly’s elbows. She stabbed a french fry, already limp with age, into a pool of ketchup. “Okay, let's run it again. Operation: Birthday Blitz. You remember the intel?” Bill, currently demolishing a stack of pancakes, mumbled around a mouthful. “Seven-year-old Timmy. Obsessed with dinosaurs. Parents loaded. Standard suburban shindig. Bounce house. Clowns. The works. And a mountain of presents.” “A *mountain*, Bill. That’s the key. Mrs. Henderson, Timmy’s mother, practically advertised the haul on Facebook. 'Timmy is SO lucky!' followed by approximately 78 photos of brightly wrapped boxes. We're talking a Nintendo Switch, a remote control T-Rex, a year’s supply of slime, the usual.” Bill swallowed the last of his pancakes. “Slimy. Excellent. Distraction potential.” “Not distraction potential, Bill. We’re not *using* the slime, we’re *stealing* everything before anyone notices it’s gone.” Molly sighed. “This is why I need to run through it. You get easily sidetracked by…shininess.” "Hey! I have focus. Especially when there's valuable merchandise involved." Bill leaned forward conspiratorially. “So, I pose as the dinosaur mascot. Full inflatable T-Rex suit. Guaranteed chaos." Molly pinched the bridge of her nose. "Absolutely not. You in a dinosaur suit is a walking disaster. Last time you tried a costume, you ended up wedged in a revolving door dressed as a giant banana." “That was a *particularly* narrow revolving door.” Bill defended himself. "And I got out eventually! Besides, kids love dinosaurs.” “They love dinosaurs that don’t trip over their own feet and accidentally stomp on a cake. You're on crowd control. Keep the horde occupied. Juggle. Make balloon animals. Be…festive. I'll handle the gift acquisition.” “Festive? Me? I’m more ‘subtle infiltration’ than ‘joyful revelry’.” “Exactly why you’re crowd control. It's a vital role. Divert attention while I discreetly…relocate the presents into the van.” Bill’s eyes lit up. “Relocate. I like that. Sounds professional. What about security?” “Two bored teenagers, supposedly hired to ‘watch things’. I’ve accounted for them. A strategically placed bag of gummy bears should keep them occupied. They’re easily bribed." "Gummy bears. Brilliant. What kind?” “Red. They’re their favourite. I staked out the last party Mrs. Henderson threw. Extensive research, Bill, extensive research. Now, focus. Timing is crucial. We go in during the pin the tail on the dinosaur game. Maximum distraction." Bill groaned. “Oh, not the tail. That’s always humiliating. I'm terrible at pinning tails.” “You are terrible at everything involving fine motor skills, Bill. Just…stay away from the blindfolds. And for the love of all that is holy, no roaring."
created_at
"2026-02-20T20:05:46.612Z"
2/20/2026, 8:05:46 PM
model
"google/gemma-3-27b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
158
moderation_latency
(null)
generation_time
21413
tokens_prompt
93
tokens_completion
768
native_tokens_prompt
75
native_tokens_completion
682
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
0.00011512
router
(null)
provider_responses
0
endpoint_id
"8f22002c-c045-446f-a1b9-9896133536b8"
id
"chatcmpl-R3cvvhaZzcyfxz0fya4l4M8e"
is_byok
false
latency
65
model_permaslug
"google/gemma-3-27b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771617946-2G88RPy4AAt78aPqhAKr"
upstream_id
"chatcmpl-R3cvvhaZzcyfxz0fya4l4M8e"
total_cost
0.00011512
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
424
0.00%
Dialogue to Total Word Ratio
Ratio: 84.43%, Deviation: 34.43%
neededClean
false
wordsTotal
424
wordsDialogue
358
0.0000%