NC Bench
Overview
About
Tests
Models
Model
Z.AI GLM 4.7 Flash
Test
Dialogue tags
Scenario
Write 500 words with 30% dialogue
Run ID
2026-02-08-01KGYZ300P829TR7CQ218S0AYJ
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 surveillance photo was stuck to the dashboard with a thick glob of chewed gum, capturing the image of a suburban tract home illuminated only by the light of a dozen glowing PlayStation consoles. Molly adjusted her fedora, tilting it so the brim cast a shadow over one eye, though the interior of the car was pitch black. "Look at the target, Bill. It’s so... innocent. So lacking in genuine security infrastructure." She gestured wildly at the windshield. "Do you see that? That is thirty unguarded tiers of human greed." "It's a birthday party," Bill grumbled, swatting a moth away from the crack. "Kids have cake. They have a bouncy castle. It’s loud." "Precisely. The noise floor provides the perfect cover for heavy fiscal extraction. And look at the tablecloth on the front porch." She pointed a gloved finger at a small speck in the photo. "Gold confetti. That’s the code." "You said we were stealing the presents." "I said we are liberating the assets. There is a semantic difference between theft and resource distribution. That’s why I spent six months online learning how to pick a combination lock without making a sound." Bill looked at the manual in his lap, then at her. "You bought a five-pound bag of Haribo gummy bears to distract the parents." "Because they have high metabolic rates, Bill! The sugar will lower their defenses and make them docile, allowing me to slip into the dining room unnoticed." "It will also turn them into screaming banshees." "Not if we use the dog." Molly opened the glove compartment and pulled out a squeaky rubber chicken. "My sources say the host’s golden retriever is terrified of poultry." "I can’t walk in there with a squeaky chicken, Molly. I have a reputation as a man of refined taste." "You have a reputation for tripping over electrical cords and asking people if they’ve seen your glasses. This is an upgrade. Now, plan B. If the dog doesn’t work, I’ll go in through the chimney. I wore form-fitting tactical leggings today." "The chimney is smaller than you are." "I’ve been doing cardio. And I have superpowers. Do you know how fast I can run when the cops are chasing me? You can’t outrun me, Bill. I’m faster." "You’re slower because you’re holding the bag of loot." "Details, details. Load the van. We have a schedule to keep, and twelve manicured toddler ponies to retrieve." She turned the key in the ignition. "Move it, Snagglepuss."
created_at
"2026-02-08T15:49:56.437427+00:00"
Sunday, February 8th, 2026 at 3:49:56 PM GMT+00:00
model
"z-ai/glm-4.7-flash-20260119"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3125
moderation_latency
(null)
generation_time
32883
tokens_prompt
61
tokens_completion
2193
native_tokens_prompt
66
native_tokens_completion
2210
native_tokens_completion_images
(null)
native_tokens_reasoning
1783
native_tokens_cached
2
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.0008885
router
(null)
provider_responses
0
id
"0a46d7fd5446177f7ee5951dd07189f6"
status
200
is_byok
false
latency
3125
endpoint_id
"6999081f-9017-4517-bbdb-47df8778c940"
provider_name
"Novita"
model_permaslug
"z-ai/glm-4.7-flash-20260119"
api_type
"completions"
id
"gen-1770565763-LVQrkpTdeFWO53NJ7cnJ"
upstream_id
"0a46d7fd5446177f7ee5951dd07189f6"
total_cost
0.0008885
cache_discount
1.2e-7
upstream_inference_cost
0
provider_name
"Novita"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0000%
Matches word count
n/a
neededClean
false
words
416
0.0000%
Dialogue to Total Word Ratio
Ratio: 72.36%, Deviation: 42.36%
neededClean
false
wordsTotal
416
wordsDialogue
301
0.0000%