NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-11-01KKFCD8ZNVR7F8W2M71GD350T
Run Details
Model:
Stealth: Hunter Alpha
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
22.7s
Cost:
$0.000000
Provider:
Stealth
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
11.79%
Dialogue to Total Word Ratio
(details)
39.9482%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
Molly popped the last piece of gum into her mouth, scanning the brightly colored invitation. “Alright, ‘Bouncy Bonanza for Brady’s Seventh.’ We hit the gift table during the superhero piñata melee. Standard.” Bill adjusted his binoculars, peering at the suburban backyard from their van. “Standard? Molly, there’s a clown. A literal, registered clown.” “Clowns are aesthetic, not tactical. Look at this floor plan.” She spread a crude crayon drawing across the dash. “Piñata break is at 1500 hours. Synchronized. That’s our window. The ‘Rejuvenation Station’—that’s the juice table—is our blind spot. We enter from the west, past the inflatable Thor.” Bill squinted. “You drew a smiley face where the parent-watchtower should be.” “Mrs. Gable. She’s a watcher, but she’s a ‘Pinot Grigio’ watcher. By piñata time, she’ll be half a bottle into ‘Sunset Blush’ and complaining about gluten. We are ghosts.” “I don’t know… the invitation says ‘Superheroes Only.’ Our outfits say ‘Mid-Nineties Grunge.’” “Pfft. Details. You’re… um… a distressed industrialist vigilante. I’m a nihilist anti-hero from a gritty reboot. It works.” Bill sighed, letting the binoculars hang. “What’s the haul, exactly? Twenty, thirty little wrapped boxes?” “Negative.” Molly’s eyes gleamed. “I consulted the mothership. The theme is ‘space.’ The kid’s into meteorites. That gift from the grandparents isn’t a toy, Bill. It’s a chunk of the Campo del Cielo. Four and a half pounds of extraterrestrial iron. Worth, conservatively, six large on the collector’s market.” “You’re robbing a seven-year-old of a space rock?” “We’re liberating it for a more… appreciative market. And as cover, we grab the flashy stuff. The noisy stuff. The light-up swords. We’re not monsters.” Bill began mentally tallying. “The piñata is our cue. Kids go feral. Parents dive for cover from candy shrapnel. We move, scoop, and egress through the side gate, into the van, gone before ‘Happy Birthday’ cake.” “Exactly. You’re on gift-table duty. I’ll create the diversion.” “What diversion?” Molly grinned. “I’m going to start a conga line.” “A conga line. At a children’s party.” “It’s disarming! It’s festive! It’ll draw every parent and their phone toward the patio. While you and the meteorite make for the exit, I’ll peel off at the inflatable slide. Simple.” Bill looked from the invitation to the chaotic backyard, now erupting as a giant, glittery piñata was hoisted onto a tree branch. “There are foam noodles everywhere. And face paint. My chinos, Molly.” “Consider it part of the job hazard. You’re not Bill, accountant, at a kid’s party. You’re ‘The Shadow,’ on a mission. Now, masks on. Remember: act natural, but slightly confused. Like you’re a dad who got the date wrong.” They pulled on plain black domino masks. Bill checked his reflection. “I look like a raccoon who’s going through something.” “Perfect. Authentic. Let’s roll. And for God’s sake, if the clown asks you to make a balloon animal, say you have an allergy.” They slipped from the van, the sound of shrieking children and the first whack of a bat echoing across the lawns, and merged toward the chaos, two very confused-looking superheroes on a most unusual crime.
created_at
"2026-03-11T21:21:26.017Z"
3/11/2026, 9:21:26 PM
model
"openrouter/hunter-alpha"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1940
moderation_latency
(null)
generation_time
22599
tokens_prompt
77
tokens_completion
1150
native_tokens_prompt
2111
native_tokens_completion
980
native_tokens_completion_images
(null)
native_tokens_reasoning
233
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
router
(null)
provider_responses
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1773264086-QCoczrEN14J3Zou5mF76"
upstream_id
"gen-1773264086-QCoczrEN14J3Zou5mF76"
total_cost
0
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Stealth"
Evaluation details
Result
Evaluator
Details
Meta Data
68.10%
Matches word count
n/a
neededClean
false
words
514
11.79%
Dialogue to Total Word Ratio
Ratio: 76.54%, Deviation: 6.54%
neededClean
false
wordsTotal
520
wordsDialogue
398
39.9482%