NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-20-01KHY829Z6MK3W6DPTTX5S8MX0
Run Details
Model:
Gemma 3 4B
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
8.1s
Cost:
$0.000056
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 30% 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 rain hammered against the corrugated iron roof of the abandoned laundromat, a relentless, slightly apologetic drumming. Molly, a woman built like a particularly sturdy brick, was meticulously constructing a miniature cardboard replica of the Carmichael residence, complete with tiny, glitter-covered Christmas lights. Bill, a man whose enthusiasm perpetually outweighed his common sense, was staring intently at a bag of gummy bears. “Seriously, Bill? Gummy bears? We’re discussing a high-stakes operation involving a battalion of clowns and a bouncy castle, and you're contemplating the existential joy of a cherry-flavored chew?” Bill popped one into his mouth with a resounding *plop*. “They’re strategic, Molly. Morale. Operation ‘Sugar Rush’ demands proper sustenance. Plus, little Timmy Carmichael is obsessed with them. Leverage.” “Leverage? You’re planning to bribe a seven-year-old with gummy bears to distract him from the chaos? That’s… remarkably low-key for you.” Molly adjusted a miniature cardboard chimney on the house. “This isn’t a bank robbery, Bill. It's a children's birthday party. The target isn't armored.” “It’s *expensive* presents, Molly! I’ve been observing. Sequined unicorns. Robotic dinosaurs. Enough miniature race cars to trigger a small coronary. We need the haul.” He gestured wildly with the bag. “Think of the resale value!” “I’m thinking of the bewildered parents and the inevitable call to Interpol,” she pointed out, meticulously gluing a tiny, plastic dollhouse onto the roof. “Okay, let’s go over the plan one more time. You handle the distraction. Balloons. Face paint. Maybe a rogue magic trick involving disappearing handkerchiefs.” “I was thinking a disappearing act myself,” Bill mumbled, chewing thoughtfully. “Like, into the party room. With a bunch of presents.” “Absolutely not. You are to remain *outside*. The objective is stealth, not theatricality. I’ll slip through the back gate. The security, as you know, is practically nonexistent – a bored teenage boy and a security camera that probably only records in black and white.” “Black and white is charming, Molly. Retro. It adds to the aesthetic. ‘A heist in sepia tones!’” "Just... focus. Do you remember the emergency exit?" “Yeah, yeah. By the… the inflatable shark. Right. What if he sees me?” “Then you’ll deploy the gummy bears. It’s instinctive. You’ve practiced.” “Practiced? I mostly just ate them.” Molly sighed, a sound like escaping steam. “Look, Bill, this needs to be precise. We need to be in and out in under fifteen minutes. No messy confrontations. No glitter explosions. Just… presents.” “Fifteen minutes? That's cutting it close. Can we at least get a sign? Like... 'Operation: Birthday Bonanza' written in icing?" “Absolutely not. Just… be stealthy. And for the love of all that is holy, don't engage with the cake.”
created_at
"2026-02-20T19:22:45.103Z"
2/20/2026, 7:22:45 PM
model
"google/gemma-3-4b-it"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
161
moderation_latency
(null)
generation_time
8017
tokens_prompt
93
tokens_completion
785
native_tokens_prompt
75
native_tokens_completion
668
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
0.00005644
router
(null)
provider_responses
0
endpoint_id
"d3de3bd4-81bc-48fb-924f-2a87b2a36e75"
id
"chatcmpl-RpBf65saeOJPkinEK3WZORXc"
is_byok
false
latency
55
model_permaslug
"google/gemma-3-4b-it"
provider_name
"DeepInfra"
status
200
api_type
"completions"
id
"gen-1771615365-QEogPgSkCIholCSUYjCC"
upstream_id
"chatcmpl-RpBf65saeOJPkinEK3WZORXc"
total_cost
0.00005644
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
443
0.00%
Dialogue to Total Word Ratio
Ratio: 74.94%, Deviation: 44.94%
neededClean
false
wordsTotal
443
wordsDialogue
332
0.0000%