NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1K5QYB8W2RQCCEFP43XAHE7
Run Details
Model:
Muse Spark 1.2 (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
58.9s
Cost:
$0.041705
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
86.38%
Matches word count
(details)
74.49%
Dialogue to Total Word Ratio
(details)
80.4331%
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 blueprint of the bouncy castle across the kitchen table and weighted the corners down with cold pizza slices, the grease soaking into the carefully drawn escape routes. Bill stared at it like it might explode and took a nervous step backward. "Okay, so walk me through it again," Bill said. "We are robbing a six-year-old?" "Not robbing, liberating," Molly said. "Timmy has thirty presents and he can't appreciate them all. It's economics." "Timmy is your nephew." "He's my second cousin once removed's kid. That's basically a stranger." Bill tapped the drawing of the backyard, his finger hovering over the side gate marked with a red X. "And your big plan is to dress as clowns?" "Genius, right? No one questions a clown at a children's party. We walk in, we juggle badly, we leave with LEGOs and a Nintendo Switch." "We don't know how to juggle." "Which is why we juggle badly. It's method acting." Bill sighed and rubbed his temples as if trying to erase the idea. "Molly, last time we tried costumes you got stuck in the air vent dressed as Santa." "And did we get the cookies? Yes. This is smaller scale. Easier." "There will be parents. And children. Children are terrifying. They stare into your soul." "Children are easily distracted. That's why we bring glitter. You throw glitter, they chase it like pigeons." Bill frowned, unconvinced, and folded his arms tightly across his chest. "You said you mapped the present table." "I did. It's by the cake. Guarded by the mother. Code name: Momzilla." "Is she dangerous?" "She organized this party with a color-coded spreadsheet. She is the most dangerous woman I've ever met." "Great. So how do we get past her?" "We wait for the singing. Everyone sings happy birthday, everyone closes their eyes for the wish, we grab the bag and go through the side gate." "What about the clown we hired? The real one?" "He canceled. Said we didn't pay his deposit. Rude." Bill leaned closer, lowering his voice to a conspiratorial whisper that barely masked his panic. "What if they ask us to make balloon animals?" "We say our specialty is invisible balloon animals. Very avant-garde. Kids love it." Molly rolled up the blueprint with a confident flick, sending a pepperoni sliding to the floor. "The presents are piled by the fence. We go in at three fifteen when the magician does his trick. Total chaos. Perfect cover." "There is a magician?" "Yeah. Uncle Dave. He does card tricks that don't work." "And if we get caught?" "We say we're performance artists teaching kids about income inequality. They'll applaud." Bill picked up a pizza slice and ate it, blueprint grease and all, chewing slowly while considering his life choices. "I still feel bad stealing from a kid." "We're not stealing from the kid. We're stealing for the kid. He doesn't need three slime factories. We'll donate one to charity. Us. We're the charity." "This is why no one invites us to family functions." "Exactly. So we have to invite ourselves."
created_at
"2026-09-03T08:21:38.801Z"
9/3/2026, 8:21:38 AM
model
"meta/muse-spark-1.2-20260805"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
27812
moderation_latency
(null)
generation_time
58824
tokens_prompt
75
tokens_completion
921
native_tokens_prompt
68
native_tokens_completion
9793
native_tokens_completion_images
0
native_tokens_reasoning
8940
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
0.04170525
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788423698-czSM0KnwUmnUIW04f9Tk"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788423698-sSprL0rtk5lNSNrhFb8J"
upstream_id
"resp_6a992e2bdbc66e2678284ac2"
provider_responses
0
endpoint_id
"ffc8e76e-808a-44d2-bd21-c6ac16e721cb"
id
"resp_6a992e2bdbc66e2678284ac2"
is_byok
false
latency
24753
model_permaslug
"meta/muse-spark-1.2-20260805"
provider_name
"Meta"
status
200
total_cost
0.04170525
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
86.38%
Matches word count
n/a
neededClean
false
words
511
74.49%
Dialogue to Total Word Ratio
Ratio: 72.43%, Deviation: 2.43%
neededClean
false
wordsTotal
515
wordsDialogue
373
80.4331%