NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYD5KSNQMFX1GZ4VRTVN2YY8
Run Details
Model:
Muse Spark 1.1 (Reasoning, Minimal)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
9.9s
Cost:
$0.005522
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.13%
Dialogue to Total Word Ratio
(details)
0.0648%
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 van smelled like old french fries and bad decisions. Molly had the blueprints spread across the dashboard. Bill was trying to untangle a single sad balloon. "Okay, so walk me through it again. We are stealing from a seven-year-old?" Bill said. "We are liberating high-value assets from a temporarily unsecured location that happens to be a children's birthday party," Molly said. "It's Timmy's party. You put on a clown nose last year and he kicked you in the shin and you cried," Bill said. "He has excellent security instincts. That's why this is a big job," Molly said. Molly tapped the hand-drawn map. It showed a house, a bouncy castle, and what was labeled 'Present Volcano' with a skull and crossbones. "3 PM. The parents will be distracted by the pinata. That's a twelve-minute window of pure chaos and blindfolded adults swinging sticks," Molly said. "And the presents? We're just going to walk in there with a sack like the world's worst Santa?" Bill said. "No. We're going in as children's entertainers. You are Mr. Wobble the Magical Clown," Molly said. "I hate clowns. Children hate me. Last time a toddler called me 'sad ham'," Bill said. "Perfect. You'll be in character. I'll be Sparkle the Face Painter. I get us past the mom perimeter," Molly said. "The mom perimeter is real. Linda has eyes everywhere. She saw me take two juice boxes last time and she still brings it up," Bill said. "Focus. The presents are in the living room. According to my intel, which is Timmy's older brother Jake, there are at least twenty unopened boxes in there. One of them is a super-rare Galactic Ranger Blaster. Worth two hundred bucks online," Molly said. "So we're committing a felony for a toy blaster?" Bill said. "It's not about the money, Bill. It's about respect. And also the money. Mostly the money," Molly said. "What's the exit strategy? We can't just sprint out carrying a pile of Paw Patrol wrapping paper. The dads will tackle us," Bill said. "We use the decoy. We bring decoy presents, swap them for the real ones. The decoy boxes will be filled with educational items. Like Brussels sprouts and math worksheets," Molly said. "That is diabolical. Those kids will be furious. Timmy will cry and then Linda will cry and then everyone will be crying," Bill said. "That's how you know it's working. While they're emotionally compromised, we wheel the real haul out in the cooler labeled 'Extra Ice'," Molly said. "Last time you labeled the cooler 'Extra Ice' you left me in a playground parking lot for three hours," Bill said. "That built character. Do you want in or out? Because I already rented the clown shoes and they don't do refunds," Molly said. Bill stared at the clown shoes. They were size twenty and squeaked when you looked at them. "If we get caught, we have to do the Bouncy Castle of Shame. Linda makes criminals bounce until they apologize," Bill said. "We won't get caught. We're professionals. What's the worst that happens at a children's party? We get cake?" Molly said. "We get put on the neighborhood watch list again. And we have to move vans again," Bill said. "Put the nose on, Bill. We're going to be the most terrifying and successful entertainers those kids have ever seen," Molly said. "I want thirty percent and the blaster," Bill said. "You get twenty percent and a juice box, and I won't tell Linda you ate all the crustless sandwiches last time," Molly said. "Deal."
created_at
"2026-07-25T17:36:50.88Z"
7/25/2026, 5:36:50 PM
model
"meta/muse-spark-1.1-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3722
moderation_latency
(null)
generation_time
9776
tokens_prompt
75
tokens_completion
876
native_tokens_prompt
225
native_tokens_completion
1233
native_tokens_completion_images
0
native_tokens_reasoning
448
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
0.0055215
router
(null)
provider_responses
0
endpoint_id
"b2b9f6f9-8880-41c1-bd0c-867650fd5238"
id
"resp_6a64f4334b5cc37addd04123"
is_byok
false
latency
335
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-1785001010-m0YpDe8kOze7avFPGeUd"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1785001010-hmB98gBaFitIjnzxoyeT"
upstream_id
"resp_6a64f4334b5cc37addd04123"
total_cost
0.0055215
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
0.00%
Matches word count
n/a
neededClean
false
words
595
0.13%
Dialogue to Total Word Ratio
Ratio: 81.53%, Deviation: 11.53%
neededClean
false
wordsTotal
601
wordsDialogue
490
0.0648%