NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD5A529J35P0VMP8G3Q6KM4
Run Details
Model:
Muse Spark 1.1 (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
23.3s
Cost:
$0.017188
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.95%
Dialogue to Total Word Ratio
(details)
99.9750%
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
The surveillance van smelled like stale french fries and terrible life choices. Molly lowered the binoculars from the Henderson house across the street. Rainbow balloons bobbed over the fence and a clown was already sweating through his makeup. Bill stabbed at a juice box. "Molly, this straw is defeating me on a spiritual level." "Bill, we have a six-year-old's birthday fortress to infiltrate and you are losing to Capri-Sun." "We need a plan." He finally punctured it. "I have a plan. We walk in, say we're the entertainment, grab the present mountain and run." "Bill, our combined entertainment value is negative. Last time you tried balloon animals you made something that looked like a divorced snake." "That snake had character. And legal grounds for alimony." "Listen. The presents are in the living room. The mom is guarding the cake like it's Fort Knox. The dad is on grill duty." Molly tapped the crudely drawn map. "So what's the distraction?" "We use the clown. We bribe him." "We can't bribe a clown, Molly. They have a strict ethical code. No balloon animals for criminals." "He's wearing a shirt that says Will Juggle For Beer. I think his ethics are flexible." Bill slurped loudly. "Okay, so I distract the parents by pretending my van broke down and I desperately need twelve juice boxes for my soccer team that is invisible." "That's the worst plan I've ever heard. And I've heard your plan to rob a bank dressed as ATMs." "It was conceptual." "My plan is we go in as superheroes. Kids love superheroes. Captain Cardboard Box and his sidekick Juice Box Boy." "You want to steal from children while dressed as their heroes? That's a new low, even for us." "Think of it as a lesson. Never leave a pile of unguarded LEGO sets near an open window." Outside, the clown's balloon popped and six children screamed. "Okay, final plan. You go in as Elsa. You sing, they are mesmerized, I steal the presents." "I am not singing Let It Go to a bunch of sugar-crazed six-year-olds. My vocal range is felony." "Fine, I'll sing. You steal. You have nimble, childlike fingers." "That is objectively true. I moisturize." "So we go on three? One, two—" The back door of the Henderson house flew open and a small girl pointed directly at the van. "Mommy! Why are those pirates living in our van?" "See? Kids already think we're awesome. Let's go rob a birthday." Molly sighed and pulled her ski mask down, which was just a wool hat with eye holes her mother had knitted. Bill put on his own mask, which was a Batman mask with the ears cut off. "Do I look intimidating or like I need PTA approval?" "You look like you pay taxes on time. Perfect cover." They stepped out of the van as the sprinkler turned on, soaking their shoes, and the clown stared at them with deep professional sympathy. Molly checked her pockets for stolen candy.
created_at
"2026-07-25T17:31:34.866Z"
7/25/2026, 5:31:34 PM
model
"meta/muse-spark-1.1-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
20409
moderation_latency
(null)
generation_time
23081
tokens_prompt
75
tokens_completion
738
native_tokens_prompt
225
native_tokens_completion
3978
native_tokens_completion_images
0
native_tokens_reasoning
3314
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
0.01718775
router
(null)
provider_responses
0
endpoint_id
"b2b9f6f9-8880-41c1-bd0c-867650fd5238"
id
"resp_6a64f2f791b1bdfe6f7b4eeb"
is_byok
false
latency
233
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-1785000694-Vm2eJtddnINB2WXE3VeU"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785000694-XPumg9DFseckQfB240lv"
upstream_id
"resp_6a64f2f791b1bdfe6f7b4eeb"
total_cost
0.01718775
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
499
99.95%
Dialogue to Total Word Ratio
Ratio: 69.90%, Deviation: 0.10%
neededClean
false
wordsTotal
505
wordsDialogue
353
99.9750%