NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-13-01M4TAJGWPR3G246HYW1SPEHF9
Run Details
Model:
Muse Wonder (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
33.7s
Cost:
$0.098412
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
96.85%
Dialogue to Total Word Ratio
(details)
98.4253%
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 spread the birthday invitation across the kitchen table while Bill leaned over her shoulder eating cereal from the box and frowning at the glitter. "Okay, listen carefully because this is the most professional job we have ever attempted and I need you fully focused," she said. "I am focused. I am also confused because this invitation has a unicorn wearing sunglasses and it says Timmy turns six on Saturday." "Exactly. Six year olds get expensive presents now. Tablets, games, bikes, those robot dogs that bark in Spanish. We go in, we take everything." "We are stealing from a six year old? That feels spiritually expensive, Molly." "We are redistributing joy from a child who has too much joy to two adults who have almost none." "That is a beautiful speech for court." Bill poured cereal onto the invitation. Molly brushed it off and pointed at the address. "Party starts at two. Parents arrive, drop presents on the gift table, then wander outside for lemonade and polite lies about parenting. That table is unattended for at least nine minutes." "Nine minutes to steal an entire mountain of toys while thirty children scream near a bouncy castle? That sounds less like a heist and more like volunteering for trauma." "Hence the disguises. You will be Bubbles the Clown. I will be your glamorous assistant, Sparkles." "Absolutely not. Last time I wore clown makeup, a goose followed me home and the makeup stayed on during my driving test." "This time is different. This clown carries a suspiciously enormous sack labeled balloon animals but actually contains stolen Lego sets." "Children will demand balloon animals. I can only make a snake, a worm, and a disappointing sausage." "Perfect. Make sausages quickly, distract them, smile through the existential dread, and drift toward the presents like a cheerful fog." "What will you do while I fog?" "I manage logistics, timing, escape routes, and the getaway tricycle if our car gets blocked by a minivan." "A tricycle cannot carry a bike, Molly. That is basic physics and also basic tricycle." She unfolded a map drawn in crayon. "Then we prioritize. Small boxes first, envelopes definitely, anything that rattles expensively second, and absolutely nothing homemade unless it looks accidentally valuable." "What about the birthday cake situation? Because if I am near cake, my moral compass melts." "No cake. Cake slows the clown. Cake creates evidence. Cake leaves frosting fingerprints on justice." Molly rolled the map, handed Bill a red nose, and watched him practice his saddest balloon sausage beside the sink. Outside, rain polished the street, their battered car coughed awake, and somewhere across town Timmy slept unaware that Saturday would bring clowns, chaos, missing robots, and two thieves arguing behind a bouncy castle. Parents would applaud what they assumed was committed entertainment, while Molly timed exits and Bill rehearsed smiling without looking like a haunted garden statue come alive. Then Bill sneezed, honked the nose, dropped his cereal, and stepped into the refrigerator door with tragic grace.
created_at
"2026-10-13T09:47:50.561Z"
10/13/2026, 9:47:50 AM
model
"meta/muse-wonder-20261012"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1915
moderation_latency
(null)
generation_time
33677
tokens_prompt
75
tokens_completion
917
native_tokens_prompt
68
native_tokens_completion
4907
native_tokens_completion_images
0
native_tokens_reasoning
4104
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
service_tier
"auto"
usage
0.098412
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791884870-wD9GzzAAzpUO28tFzjOH"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791884870-lxGd46Sn3cB1c29yqsFO"
upstream_id
"resp_6acdfe47162e190b16094f10"
provider_responses
0
endpoint_id
"1232618d-ebfe-4338-9b68-d9127a57c8dd"
id
"resp_6acdfe47162e190b16094f10"
is_byok
false
latency
588
model_permaslug
"meta/muse-wonder-20261012"
provider_name
"Meta"
status
200
total_cost
0.098412
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Meta"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
96.85%
Dialogue to Total Word Ratio
Ratio: 70.80%, Deviation: 0.80%
neededClean
false
wordsTotal
500
wordsDialogue
354
98.4253%