NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1GTD4CGAJYVW1DA99QXWRCF
Run Details
Model:
Z.AI GLM 5.3 Flash (Reasoning, Max)
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
1m 11s
Cost:
$0.000816
Provider:
Z.AI
Evaluation summary
Score
Evaluator
Details Link
97.63%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
98.8138%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 50% 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 unrolled the blueprint across the diner table. It was a crayon drawing of a backyard, complete with a smiling sun. "Target's the gift table," she said, tapping a cluster of green scribbles. "Bouncy castle here. Clown patrol there." Bill leaned away like the paper had bitten him. "I don't do clowns, Molly. You know this." "You did the panda job in '09." "Pandas have the decency to be real bears." She rolled her eyes and nudged a plastic dinosaur two inches across the table. "Fine. I'll handle Chuckles. You take the piƱata. When it bursts, every kid in a thirty-foot radius dives for candy. Perfect chaos window." "And the parents?" Mothers, Molly explained, were the real security system. Watch for the one radiating walkie-talkie energy. Her name would be Karen. It was always Karen. Bill traced the fence with one finger. "Exit strategy?" "Bounce. Launch yourself over the fence from the castle." "That's not an exit. That's a lawsuit." Molly just grinned. They'd strike at three sharp, mid-cake-song, when nobody guarded presents. She stood, pocketing the blueprint. "Wear the dinosaur costume." Bill stared. "Six-year-olds chased me two blocks last time." "They're like tiny drunk cops. Adorable." "Terrifying." "Three o'clock, Bill. Bring the sack. Leave your dignity."
created_at
"2026-09-02T10:25:01.338Z"
9/2/2026, 10:25:01 AM
model
"z-ai/glm-5.3-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
6930
moderation_latency
(null)
generation_time
71208
tokens_prompt
75
tokens_completion
3108
native_tokens_prompt
73
native_tokens_completion
3259
native_tokens_completion_images
(null)
native_tokens_reasoning
2954
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
0.000816385
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788344701-d1mCOuzxkDpbKSbPrPb6"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788344701-quEYSb6xLjg6cE0XFT8E"
upstream_id
"2026090218250165ad6c51eded4524"
provider_responses
0
endpoint_id
"8e9fe48b-2f91-41c3-a8a7-e4a93a8c4ff0"
id
"2026090218250165ad6c51eded4524"
is_byok
false
latency
6930
model_permaslug
"z-ai/glm-5.3-flash-20260826"
provider_name
"Z.AI"
status
200
total_cost
0.000816385
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
97.63%
Matches word count
n/a
neededClean
false
words
207
100.00%
Dialogue to Total Word Ratio
Ratio: 50.23%, Deviation: 0.23%
neededClean
false
wordsTotal
213
wordsDialogue
107
98.8138%