NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1K4QV2MAFTT2JC9TRVC4WHS
Run Details
Model:
Muse Spark 1.2 (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
42.9s
Cost:
$0.039487
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
97.63%
Matches word count
(details)
48.58%
Dialogue to Total Word Ratio
(details)
73.1028%
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 crouched behind the hedge staring at the bouncy castle while Bill tried to untangle himself from the neighbor's recycling bin. The yard beyond was pure chaos, screaming kids, a wheezing clown, and a tall pile of presents glittering in the afternoon sun. "Okay, walk me through the plan again, Bill, because I still think you are insane." "Insane is a strong word. I prefer creatively ambitious." "You want to rob a six-year-old's birthday party. That is not ambitious, that is morally bankrupt and also embarrassing." "It's not just any six-year-old. It's Timmy Henderson. His dad is Henderson the toy magnate. The presents are not socks and coloring books. We are talking limited edition robots, solid gold Legos, the works." "So your big score is stealing from children? Have you considered just getting a job?" "Have you considered supporting my dreams? Look, the backyard is wide open. The clown is distracting everyone, the parents are distracted. We go in dressed as entertainers, grab the presents, and vanish. Clean, simple, victimless." "Victimless? The victims are literally children. They will cry. Loudly. Children are professional criers." "They'll get over it. Kids are resilient. Plus we will leave the cake." "Oh, how generous to leave the cake. You are a saint among thieves." "Thank you. Now, you are the magician and I am the balloon guy. I have practiced." "You can't make a balloon animal. Last time you made something that looked like a sad intestine." "That was an avant-garde poodle. This time I have a manual. Look, I can make a sword. Watch." Molly watched Bill twist the balloon until it popped loudly, causing three children to stare in their direction. The clown stopped mid-juggle and squinted toward the hedge. "Great. You just announced our position. The toddlers have spotted us." "That was a tactical pop. To test their reflexes. They failed." "They are four, Bill." "Exactly. Easy marks. So do we go now? The present table is right there. I can see a box wrapped in dinosaur paper that is literally shaking. Something expensive is in there." Bill bounced on his toes, vibrating with nervous energy as his balloon pump squeaked. He grinned under his crooked red nose, already picturing gold bricks. "We are not stealing shaking dinosaur presents from toddlers. We could steal the clown's car instead." "The clown's car is a 1998 Honda Civic painted like a shoe. No thanks. The presents are where the value is." "Fine. But if we get caught, I'm telling the cops you kidnapped me and forced me to wear this cape." "Deal. And if we succeed, you get the robot and I get the gold Legos. And we never speak of this again." Molly sighed and pulled the cape tighter, resigning herself to the stupidest crime of her career as the party music started again. The bouncy castle wheezed behind them as parents sang badly off-key nearby. "Agreed. Let's go rob the worst heist in history."
created_at
"2026-09-03T08:04:06.875Z"
9/3/2026, 8:04:06 AM
model
"meta/muse-spark-1.2-20260805"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2175
moderation_latency
(null)
generation_time
42892
tokens_prompt
75
tokens_completion
751
native_tokens_prompt
68
native_tokens_completion
9271
native_tokens_completion_images
0
native_tokens_reasoning
8584
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
0.03948675
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788422646-SMGMGUyg9yG39d9YqQuG"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788422646-pyx2yAsTUutcF9ZiN8JF"
upstream_id
"resp_6a9929f77eae25cb0f5149d6"
provider_responses
0
endpoint_id
"ffc8e76e-808a-44d2-bd21-c6ac16e721cb"
id
"resp_6a9929f77eae25cb0f5149d6"
is_byok
false
latency
510
model_permaslug
"meta/muse-spark-1.2-20260805"
provider_name
"Meta"
status
200
total_cost
0.03948675
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
97.63%
Matches word count
n/a
neededClean
false
words
493
48.58%
Dialogue to Total Word Ratio
Ratio: 73.80%, Deviation: 3.80%
neededClean
false
wordsTotal
500
wordsDialogue
369
73.1028%