NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KRM700FFB0VWKQZ95QM6CT
Run Details
Model:
Muse Glimmer 30B (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
32.4s
Cost:
$0.002240
Provider:
Phala
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
0.0000%
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 rain had turned the cul-de-sac into a mirror and Molly was already crouching on the hedgerow with a juice box in one hand and a clipboard in the other. Bill was beside her in a rented Hello Kitty costume that smelled faintly of dry cleaning and poor decisions. The Hendersons' backyard was lit up like a birthday bomb, balloons bobbing, a plastic slide glowing pink, and a table sagging under the weight of presents wrapped in enough glitter to blind a small animal. Molly lowered her binoculars. The target was six years old, a boy named Tyler, and his party was in full swing with a piñata shaped like a dinosaur and a mother who kept refilling a punch bowl. The presents were stacked in a neat row against the fence, untouched until cake. The whole yard smelled like hot dogs and wet grass. Bill tugged at his cat ears. This is insane, she said. This is a children's party. Molly didn't look up. That's why it's perfect. No security, no alarms, just sugar and a mother who thinks glitter is a food group. Bill sighed. We steal from museums, banks, sometimes art galleries. We don't steal birthday presents. We do now, Molly said. Client wants unopened toys. Wants them before the kid opens them. Wants the dopamine intact. That's weird. That's the job. She flipped her clipboard and tapped a diagram drawn in crayon. Entry point is the back gate, left unlocked since Mrs. Henderson always forgets. We go in during the piñata swing, everyone is distracted, kids scream, parents cheer. We grab the presents, dump them in the wagon, out the gate in ninety seconds. What about the clowns? Bill asked, pointing at a man in a rainbow wig handing out cupcakes. That's not a clown, that's Kevin from accounting. He got voluntold. He's harmless unless you mention his divorce. Molly grinned. Harmless is good. We wear the costumes, we blend. Bill frowned at his own reflection in a car window. I look like a crime. You look like a mom who lost a bet, Molly said. Flattering. The music swelled. A kid shouted about a dinosaur. Bill checked his watch. Thirty seconds to go. Do we have a cover story if we get caught? Molly was already standing, adjusting the oversized bow on her head. We are the late gift delivery. Very late. Very apologetic. If anyone asks, we're from Party City and we got the wrong address. Perfect. Bill whispered. Please tell me the client isn't a kidnapper. Client is a collector, Molly said. Collects disappointment. He buys unopened presents on the dark web and resells them to adults who never got one. It's a whole market. That's the saddest thing I've ever heard. Agreed. Let's go. They slipped through the gate as the piñata finally cracked open, a cascade of candy raining down like confetti. Parents clapped and phones came out. Molly moved first, scooping boxes with both arms while Bill kept the wagon rolling and humming the party song off key. A toddler tried to hug his leg and Bill froze. It's okay, honey, he said softly, I'm just here for the presents. The mother laughed and offered him a cupcake. He took it to be polite. They were out the gate in eighty seven seconds with twelve boxes, one inflatable dinosaur, and a suspicious amount of confetti in their hair. In the car Bill exhaled hard. We just robbed a six year old. Molly was already checking the boxes by flashlight, peeling back tape. We liberated them. And for once, nobody cried.
created_at
"2026-09-03T13:51:39.53Z"
9/3/2026, 1:51:39 PM
model
"meta/muse-glimmer-30b-20260810"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
266
moderation_latency
(null)
generation_time
32304
tokens_prompt
75
tokens_completion
2295
native_tokens_prompt
118
native_tokens_completion
2032
native_tokens_completion_images
(null)
native_tokens_reasoning
1261
native_tokens_cached
117
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
0.00224018
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788443499-PTUGOBdC3fAKUGLtDTU2"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788443499-if7tC7MRkOQf5HTnGqM0"
upstream_id
"req_7acbd5e2cfd0dcf698be9797c02ee2a5"
provider_responses
0
endpoint_id
"6244f10e-1e45-488b-a47f-f575e0edd12b"
id
"req_7acbd5e2cfd0dcf698be9797c02ee2a5"
is_byok
false
latency
225
model_permaslug
"meta/muse-glimmer-30b-20260810"
provider_name
"Phala"
status
200
total_cost
0.00224018
cache_discount
0.00003042
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
0.00%
Matches word count
n/a
neededClean
false
words
600
0.00%
Dialogue to Total Word Ratio
Ratio: 5.65%, Deviation: 24.35%
neededClean
false
wordsTotal
602
wordsDialogue
34
0.0000%