NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHYAXQDKP6FBX4V88CVQ40G5
Run Details
Model:
Gemma 3 12B
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
11.5s
Cost:
$0.000086
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 flickering neon sign of “Bob’s Bait & Tackle” cast a sickly green glow across the interior of the van. Molly chewed on a pencil, tracing circles on a crumpled blueprint of “Timothy Higgins’ Fifth Birthday Bash – A Pirate Adventure!” Bill, meanwhile, was meticulously arranging a collection of rubber chickens on the dashboard. “Rubber chickens, Bill? Seriously?” Molly finally said, breaking the silence. “We’re stealing gifts, not staging a poultry-themed performance art piece.” Bill didn't look up. "They’re for distraction, Molly. Think about it. A sudden, coordinated rubber chicken release amidst a horde of sugar-fueled five-year-olds. Chaos. Diversion. Opportunity." He punctuated his explanation with a dramatic flourish. Molly sighed, running a hand through her already messy bun. “You know, sometimes I wonder if you actually understand the word ‘heist.’” “Of course I do! It’s about audaciousness, Molly! About thinking outside the toy box! About...rubber chickens.” He held one up, admiring it. “This little guy is particularly menacing, don’t you think?" "It's rubber, Bill." “Precisely! Unexpected! Now, about the plan. Security is minimal. One bored-looking teenager supposedly guarding the bouncy castle. Our primary target is the gift table – a veritable treasure trove of plastic weaponry, princess dresses, and suspiciously cheap Lego sets.” “Right. Bouncy castle guard. Easily dealt with. You think you can hypnotize him with a rubber chicken?” Bill shrugged. “Worth a shot. Plan B is offering him a particularly shiny one. People are surprisingly susceptible to shiny things.” “Plan C is simply… talking to him. Straightforward. Minimal poultry involvement.” "Where's the fun in that, Molly? Where's the *theatricality*?” He gestured widely. "Remember the museum job? The synchronized flamingo dance? That was a masterpiece!" Molly shuddered. “Let's not revisit the flamingo incident, please. The curator still sends me hate mail.” "Minor setback. Think of the media coverage! Gentle thieves, wielding fowl, liberating joy!" “We're stealing presents, Bill. Not liberating joy. Timothy Higgins’ parents specifically requested no clowns. A rubber chicken invasion is probably going to push them over the edge.” “Okay, okay, less chicken. More… strategic snatches. You handle the bouncy castle guard. I’ll coordinate the gift acquisition. I’ve already calculated the optimal route to ensure maximum present-per-minute efficiency.” He tapped the blueprint with a surprisingly serious expression. "See? This section, beneath the pirate ship piñata, is particularly promising. Lots of potential for high-value items.” Molly leaned closer, squinting at the drawing. “Are those… miniature ponies?” “Among other delights! Imagine the resale value on those, Molly! Miniature ponies! We’ll be set!” “I just hope those ponies don’t bite.”
created_at
"2026-02-20T20:12:40.766Z"
2/20/2026, 8:12:40 PM
model
"google/gemma-3-12b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
189
moderation_latency
(null)
generation_time
11301
tokens_prompt
93
tokens_completion
788
native_tokens_prompt
75
native_tokens_completion
640
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
0.0000862
router
(null)
provider_responses
0
endpoint_id
"eb06dc92-5a16-47ec-a776-6ef956457c47"
id
"chatcmpl-RlmwhVegrFTrewtxOBRSPZvV"
is_byok
false
latency
115
model_permaslug
"google/gemma-3-12b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771618360-o4mXA2aOein1gBBSXAUg"
upstream_id
"chatcmpl-RlmwhVegrFTrewtxOBRSPZvV"
total_cost
0.0000862
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
426
0.00%
Dialogue to Total Word Ratio
Ratio: 77.00%, Deviation: 27.00%
neededClean
false
wordsTotal
426
wordsDialogue
328
0.0000%