NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1GWCRYSCY6VK0DSV04T4M1D
Run Details
Model:
Z.AI GLM 5.3 Flash (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
3m 32s
Cost:
$0.002555
Provider:
Z.AI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
75.96%
Dialogue to Total Word Ratio
(details)
87.9792%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 crouched in the hydrangeas, binoculars perched atop her reading glasses, watching the Pembertons' backyard glow with fairy lights and the ominous hum of a rented bouncy castle. Bill rolled off a wheelie bin and landed beside her with the grace of falling furniture. "It's a birthday party, Molly." He gestured at the scene: thirty seven-year-olds orbiting a gift table stacked with wrapped offerings taller than the birthday boy himself. Somewhere, a parent fired up a fog machine, because apparently espionage aesthetics were included in the deluxe package. "With a gift table, Bill." Molly unrolled a laminated blueprint across her knees. The infiltration route was color-coded. The escape route was, concerningly, drawn entirely in crayon. "Surrounded by children." He wasn't wrong. The perimeter bristled with juice boxes. A piƱata swung overhead, a sugar-filled security system, and a grandmother patrolled the snack station like a prison guard. "Distracted children." "Eating cake." She tapped the blueprint. Phase one: enter as the clown. Phase two: swap the gift bags with decoys from Bill's garage. Phase three: exit through the moonbounce. Bill stared at her. "That's somehow worse." Molly adjusted her clown nose and smiled like a woman about to commit present larceny.
created_at
"2026-09-02T10:59:46.787Z"
9/2/2026, 10:59:46 AM
model
"z-ai/glm-5.3-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2861
moderation_latency
(null)
generation_time
212322
tokens_prompt
75
tokens_completion
9233
native_tokens_prompt
73
native_tokens_completion
10212
native_tokens_completion_images
(null)
native_tokens_reasoning
9931
native_tokens_cached
64
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
0.002554635
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788346786-7gAoWxXz82w0NN2WCM8Y"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788346786-ia7ppvzeyMH7GJ9US5h7"
upstream_id
"2026090218594768401fdfb1fa4d59"
provider_responses
0
endpoint_id
"8e9fe48b-2f91-41c3-a8a7-e4a93a8c4ff0"
id
"2026090218594768401fdfb1fa4d59"
is_byok
false
latency
2861
model_permaslug
"z-ai/glm-5.3-flash-20260826"
provider_name
"Z.AI"
status
200
total_cost
0.002554635
cache_discount
0.00000768
upstream_inference_cost
0
provider_name
"Z.AI"
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
199
75.96%
Dialogue to Total Word Ratio
Ratio: 17.24%, Deviation: 7.24%
neededClean
false
wordsTotal
203
wordsDialogue
35
87.9792%