NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ015KSCGSNS382A8C07E7J9
Run Details
Model:
Z.AI GLM 4.5 Air
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
20.1s
Cost:
$0.000669
Provider:
Z.AI
Evaluation summary
Score
Evaluator
Details Link
3.62%
Matches word count
(details)
99.86%
Dialogue to Total Word Ratio
(details)
51.7421%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
"Alright, Molly, the layout?" Bill whispers, peering through the hedge. "Clowns on the left, bouncy castle on the right, tables piled high with loot right through those big glass doors," Molly replies, sketching in the dirt. "Security?" "Dad over there? He's busy refilling the piñata," Bill grins. "And the kids are glued to the magician, Sparkles the Magnificent." "Sparkles looks like he just escaped a retirement community," Molly mutters. "Plan?" "Simple, sweetheart. You create a diversion near the punch bowl – maybe 'accidentally' knock it over?" "Bill, that's genius. Pure chaos. You grab the loot while they're mopping up." "Works for me. Meet you at the van in five minutes? Full load." "Five minutes? Bill, those presents are wrapped tighter than a Burrito Supreme! We'll need ten, maybe fifteen!" "Ten minutes, then. Don't get distracted by the cake." "Me? Distracted? Never. Now, where did I put my crowbar?" Bill asks, patting his pockets. "It's in your back pocket, idiot. You sat on it during the clown act." "Ah. Explains the wedgie. Let's go steal Christmas in July!"
created_at
"2026-04-24T15:19:11.416Z"
4/24/2026, 3:19:11 PM
model
"z-ai/glm-4.5-air"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1965
moderation_latency
(null)
generation_time
9883
tokens_prompt
77
tokens_completion
655
native_tokens_prompt
66
native_tokens_completion
597
native_tokens_completion_images
(null)
native_tokens_reasoning
373
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
0.00066871
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
"20260424231921a21d23e1c0244ef9"
is_byok
false
latency
1964
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-1777043951-OeDav0d8zrk2uqUS2R5n"
session_id
(null)
api_type
"completions"
id
"gen-1777043951-Eh8oQtzWi9tieQwgcB8n"
upstream_id
"20260424231921a21d23e1c0244ef9"
total_cost
0.00066871
cache_discount
0.00000119
upstream_inference_cost
0
provider_name
"Z.AI"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
3.62%
Matches word count
n/a
neededClean
false
words
176
99.86%
Dialogue to Total Word Ratio
Ratio: 88.07%, Deviation: 1.93%
neededClean
false
wordsTotal
176
wordsDialogue
155
51.7421%