NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD4Q1PCGHMVBZEKFXQSS7EQ
Run Details
Model:
Muse Spark 1.1 (Reasoning, Minimal)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
26.5s
Cost:
$0.019232
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
96.11%
Dialogue to Total Word Ratio
(details)
98.0538%
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 van smelled of old glitter and desperation, parked two houses down from the Hendersons' suburban fortress. Molly adjusted her too-tight rabbit costume ears while Bill tried to fold a stolen street map into something resembling competence. "Bill, remind me why we're stealing from an eight-year-old?" Inside the backyard, pastel balloons strained against the fence and a mountain of presents teetered beside a cake shaped like a dinosaur roaring. A banner screamed HAPPY BIRTHDAY LEO in Comic Sans, which offended Molly more than the actual felony. "Because, Molly, Jake the Fence said the kid's dad collects vintage comics and wraps them as presents." Molly stared at the target, a pile wrapped in superhero paper that probably contained more sugar than value. Her professional pride itched. She preferred banks, museums, at least a jewelry store with air conditioning. "So we're risking prison for potentially sticky copies of Spider-Man?" Bill, meanwhile, was applying face paint with the focus of a surgeon and the skill of a toddler. He had already painted a lopsided star on his cheek and declared himself ready for infiltration. "No, we're risking a clown encounter. I heard they hired Mr. Giggles." "You fear clowns more than the cops?" Their plan involved entry through the side gate, misdirection via terrible magic tricks, and egress with the largest bag they could carry without looking like the Grinch. Molly doubted everything, especially the fake mustache Bill wore upside down. "Cops don't twist balloons into dachshunds while maintaining eye contact." The music from the party thumped with children's laughter and a slightly unhinged entertainer shouting about pinatas. Molly could see a mother arranging juice boxes like ammunition, vigilant, terrifying, the true security system. "Fine. What's the plan, genius?" Bill pulled out a collapsible top hat that collapsed permanently. Molly sighed, the sigh of a woman who had once cracked a safe with a paperclip and now contemplated stealing Paw Patrol wrapping paper. "We dress as entertainers. You be the magician, I be your lovely assistant." "Last time I was lovely assistant, you sawed me in half and left me." She checked her watch. The cake ceremony was imminent, meaning all parental attention would pivot. That was their window, narrow as Bill's understanding of legal boundaries, but potentially sufficient for grabbing cardboard-covered disappointment. "That was method acting. This time, we grab the sack and vanish." Molly zipped her duffel bag, resigned. The universe had reduced her career to this, a clown, a bouncy castle, and ethically questionable loot. Still, vintage comics might pay rent, if they survived. "There's thirty kids, one bouncy castle, and zero exits. How do we vanish?" "We use the oldest trick. We cry until someone's mom gives us juice boxes and lets us go." They opened the van doors together, stepping into bright daylight, absurdly costumed, utterly unprepared, yet weirdly optimistic about crime and children's parties. "Molly, that's pathetic. Even for you." Somehow, it felt like the perfect beginning. "Watch and learn." At least until the clown arrived. "Whatever."
created_at
"2026-07-25T17:21:08.82Z"
7/25/2026, 5:21:08 PM
model
"meta/muse-spark-1.1-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
23615
moderation_latency
(null)
generation_time
26425
tokens_prompt
75
tokens_completion
814
native_tokens_prompt
225
native_tokens_completion
4459
native_tokens_completion_images
0
native_tokens_reasoning
3781
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
0.019232
router
(null)
provider_responses
0
endpoint_id
"b2b9f6f9-8880-41c1-bd0c-867650fd5238"
id
"resp_6a64f085e3a81c53aec64b2b"
is_byok
false
latency
329
model_permaslug
"meta/muse-spark-1.1-20260709"
provider_name
"Meta"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1785000068-flct2iQuJ9ybV7tYh3XA"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785000068-ubHpdv4oPJIBJIPEx8zW"
upstream_id
"resp_6a64f085e3a81c53aec64b2b"
total_cost
0.019232
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Meta"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
96.11%
Dialogue to Total Word Ratio
Ratio: 30.89%, Deviation: 0.89%
neededClean
false
wordsTotal
505
wordsDialogue
156
98.0538%