NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ05STSMD613XYMAEF0P9XEQ
Run Details
Model:
Z.AI GLM 4.5 Air
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
33.7s
Cost:
$0.000940
Provider:
Z.AI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
1.35%
Dialogue to Total Word Ratio
(details)
0.6767%
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
Molly paced, checking her watch. "Bill, you’re late. The party starts in ten minutes." Bill burst through the door, grinning, "Sorry, traffic! Got held up at the clown college. Turns out juggling chainsaws is harder than it looks." Molly rolled her eyes. "Focus. The target is the birthday boy’s room. Second floor, left window, slightlyajar." "Easy peasy," Bill rubbed his hands together. "I’ll shimmy up, unlock it, you toss me the grabber." Molly handed him a extendable claw. "And the silent alarms?" "Disabled. I bribed the clown, remember? He’s supposed to be leading the kids in a ‘musical chairs with whoopee cushions’ distraction downstairs. Loud, chaotic, perfect cover." Molly nodded. "Good. You go first. I’ll follow two minutes later, acting like a confused aunt looking for the bathroom." Bill beamed. "Brilliant! My disguise is flawless." He patted a terrible polyester jacket. "Who suspects a man wearing a neon green wig and mismatched socks?" Molly sighed. "Just don’t trip over the bouncy castle entrance on the way to the house." "Me? Trip?" Bill scoffed. "I’m as graceful as a cat on a hot tin roof. Speaking of cats..." "Bill, no." "Just one quick peek!" Bill insisted. "What if there’s a kitten gift? Can’t leave a kitten behind. That’s bad karma." Molly pinched the bridge of her nose. "Fine. One peek. Then presents. Only presents. We grab the big ones first— Lego sets, gaming consoles. Leave the dollar-store trinkets. Less conspicuous." "Got it. Prioritize the loot." Bill started towards the door, then paused. "Molly?" "What?" "...What if the birthday boy sees me?" Molly sighed again. "You’re wearing a ridiculous neon wig and clown makeup, Bill. He’ll think you’re part of the entertainment. Now move! Before the clown runs out of whoopee cushions." She pushed him gently. "And for heaven's sake, be quiet on the stairs!"
created_at
"2026-04-24T16:40:08.256Z"
4/24/2026, 4:40:08 PM
model
"z-ai/glm-4.5-air"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2085
moderation_latency
(null)
generation_time
23519
tokens_prompt
77
tokens_completion
966
native_tokens_prompt
66
native_tokens_completion
844
native_tokens_completion_images
(null)
native_tokens_reasoning
462
native_tokens_cached
7
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"
usage
0.00094041
router
(null)
provider_responses
0
endpoint_id
"f7dc79ca-bbbd-41a1-8eef-616d981ff1cb"
is_byok
false
latency
10077
model_permaslug
"z-ai/glm-4.5-air"
provider_name
"SiliconFlow"
status
504
1
endpoint_id
"c8760514-8a81-4a73-ae2c-be4942ba93f3"
id
"2026042500401891d205dd4b214719"
is_byok
false
latency
2085
model_permaslug
"z-ai/glm-4.5-air"
provider_name
"Z.AI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777048808-kqDz0xOIW7ZWQtNO6DP1"
session_id
(null)
api_type
"completions"
id
"gen-1777048808-XBW5kB09btkowOMj0MYM"
upstream_id
"2026042500401891d205dd4b214719"
total_cost
0.00094041
cache_discount
0.00000119
upstream_inference_cost
0
provider_name
"Z.AI"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
303
1.35%
Dialogue to Total Word Ratio
Ratio: 79.28%, Deviation: 9.28%
neededClean
false
wordsTotal
304
wordsDialogue
241
0.6767%