NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KR853F165F4ZKA26TXGAQX
Run Details
Model:
Muse Glimmer 30B (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
18.6s
Cost:
$0.001326
Provider:
Phala
Evaluation summary
Score
Evaluator
Details Link
68.10%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
34.0511%
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
Molly spread the flyer across the kitchen table like it was a blueprint for a bank vault. The paper showed a bright pink unicorn, a six year old birthday girl named Tilly, and a two o'clock party at the community center on Maple. Bill peered over her shoulder with a half eaten donut in his hand. We are not stealing a bank, we are stealing a birthday, she said. She tapped the flyer with a red pen. Presents by the door, cake at three, parents watching the door like it's Fort Knox. Bill wiped sugar off his lip. That's the best part. Parents watch the door. Kids watch the cake. Nobody watches the presents. Molly grinned. Exactly. The presents are just sitting there. Unlocked. Unarmed. Unsupervised except by a seven year old in a dinosaur costume. That's the risk. The dinosaur might bite, he said. Molly rolled her eyes. Bill, it's a plastic costume. The only thing dangerous here is the sugar crash. She pulled out a second sheet, a hand drawn map of the community center with an arrow to the side entrance and a note about the bathroom window. Plan is in. We go in as late guests. You be the clumsy uncle who brought the wrong present. I'll be the aunt who needs the restroom. Bill frowned at the map. And how do we get all the presents out without a van full of loot? We don't. We get the good ones. Molly pointed to the flyer. Tilly's list says a scooter, a tablet, and a build it yourself volcano. That's three trips in the diaper bag. You have a diaper bag, he said. I have a diaper bag. It's roomy, padded, and no one questions a diaper bag at a kids party. She unzipped it to show a lining of black trash bags. In and out in ninety seconds. The parents are too busy making sure the kids don't eat frosting off the floor. What about the music? he asked. There's a DJ. It's a playlist on a phone. Molly said. I checked. The dad is running it from the kitchen. Kitchen is our blind spot. Bill looked at the flyer again. Two o'clock start. We hit at 2:45. Presents are piled, cake is out, attention is divided. Perfect timing, Molly said. You grab the scooter, I'll grab the tablet. Volcano can wait for the next heist. You make it sound romantic. Bill said. It's a children's party. It's the softest target we've ever had. She folded the flyer carefully. And if anyone asks, we were just there for the cake. Bill sighed and stood up. Fine. But if a child cries because we took her volcano, I'm blaming you. Molly was already putting on her fake glasses. Deal. Just remember, Bill, you're the clumsy uncle. Trip over the rug. Knock over a juice box. Distraction is our friend. And you are the aunt who needs the restroom very badly, he said. I'll be the aunt who needs the restroom and the presents. She smiled. Let's go steal some joy.
created_at
"2026-09-03T13:45:04.374Z"
9/3/2026, 1:45:04 PM
model
"meta/muse-glimmer-30b-20260810"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
343
moderation_latency
(null)
generation_time
18540
tokens_prompt
75
tokens_completion
1299
native_tokens_prompt
118
native_tokens_completion
1186
native_tokens_completion_images
(null)
native_tokens_reasoning
524
native_tokens_cached
54
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
"stop"
service_tier
(null)
usage
0.00132596
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788443104-RefPLMnzFBY5sQe1njPr"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788443104-vaJzx1qhnRVjod3PSx1m"
upstream_id
"req_349246a46aed444143c609d592a79546"
provider_responses
0
endpoint_id
"6244f10e-1e45-488b-a47f-f575e0edd12b"
id
"req_349246a46aed444143c609d592a79546"
is_byok
false
latency
304
model_permaslug
"meta/muse-glimmer-30b-20260810"
provider_name
"Phala"
status
200
total_cost
0.00132596
cache_discount
0.00001404
upstream_inference_cost
0
provider_name
"Phala"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
68.10%
Matches word count
n/a
neededClean
false
words
514
0.00%
Dialogue to Total Word Ratio
Ratio: 0.00%, Deviation: 30.00%
neededClean
false
wordsTotal
514
wordsDialogue
0
34.0511%