NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1K61S3K9Y76A8D2XAXN5X09
Run Details
Model:
Muse Spark 1.2 (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
18.4s
Cost:
$0.022032
Provider:
Meta
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.76%
Dialogue to Total Word Ratio
(details)
99.8806%
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 faintly of old coffee and stolen ambition. Molly spread a crayon-drawn map across the dashboard, which she had confiscated from the target's older brother for two dollars and a juice box. The map showed a suburban backyard rendered in enthusiastic purple marker, complete with a lopsided castle, a rectangle labeled PRESENTS!!! and a stick figure clown that looked deeply menacing. Bill stared at it like it was the blueprint to Fort Knox, except Fort Knox had fewer juice boxes. Outside the elementary school, minivans were already lining up and mothers were unloading platters of cupcakes that could feed a small army. "Bill, we are stealing from a six-year-old? Have we finally hit rock bottom?" "Molly, rock bottom is stealing candy from a baby. This is presents from toddlers. It's aspirational." Molly tapped the present table with a chipped pink nail. According to her intel, the birthday boy, Timmy, was turning six and had specifically requested no socks, which meant every parent had overcompensated with large, expensive, brightly wrapped boxes. The haul would be tremendous, if morally bankrupt. Bill was adjusting his disguise, which consisted of a stolen entertainer's polo shirt that was three sizes too small and a name tag that read HELLO I AM FUN. "The invite said every kid brings a gift. That's twenty gifts, Bill. Twenty untouched action figures and glitter ponies." "And a bounce house. Don't forget we have to get past the bounce house." "That's not security, Bill. That's inflatable joy." The plan required surgical precision, or at least the kind of precision two adults who had once failed to steal a hot dog stand could manage. Molly had parked two blocks away to avoid suspicion, although nothing says suspicion quite like two adults in a van giggling at a child's party invitation. She produced a pair of oversized sunglasses and a bag labeled TOTALLY NOT FOR STEALING PRESENTS. "Tell that to my sciatica. Last time I went in a bounce house I couldn't walk for three days." "Okay, plan is simple. You distract the clown, I grab the gift table and run." "Why do I always get the clown? Last time the clown cried." The clown was apparently named Mr. Chuckles and had a reputation for making balloon animals that looked like existential dread. Bill shuddered. Somewhere in the distance, children screamed with delight, which Molly chose to interpret as the sound of opportunity. She folded the map with the reverence of a general preparing for battle, then checked her watch. The party started in ten minutes. "Because you have trustworthy clown energy. I have suspicious ponytail energy." "Fine. But if I get stuck in the ball pit again, you owe me pizza forever." Both thieves took a deep breath, feeling ridiculous and oddly professional, as if robbing a bounce house required the same gravitas as robbing a bank vault. Molly grinned, sensing victory smelled suspiciously like frosting and desperation. It was perfect. "Deal. Now let's steal childhood innocence." "Let's go."
created_at
"2026-09-03T08:27:01.12Z"
9/3/2026, 8:27:01 AM
model
"meta/muse-spark-1.2-20260805"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2549
moderation_latency
(null)
generation_time
18316
tokens_prompt
75
tokens_completion
897
native_tokens_prompt
68
native_tokens_completion
5164
native_tokens_completion_images
0
native_tokens_reasoning
4387
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
0.022032
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788424021-KQTLgFrwEZMUkqHHKluV"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788424021-tu2ltLENzcODPValypwE"
upstream_id
"resp_6a992f552e65a98967cc4d15"
provider_responses
0
endpoint_id
"ffc8e76e-808a-44d2-bd21-c6ac16e721cb"
id
"resp_6a992f552e65a98967cc4d15"
is_byok
false
latency
1074
model_permaslug
"meta/muse-spark-1.2-20260805"
provider_name
"Meta"
status
200
total_cost
0.022032
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
100.00%
Matches word count
n/a
neededClean
false
words
500
99.76%
Dialogue to Total Word Ratio
Ratio: 30.22%, Deviation: 0.22%
neededClean
false
wordsTotal
503
wordsDialogue
152
99.8806%