NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ080WAFRTDVAWN6HDPXTNG5
Run Details
Model:
Z.AI GLM 5.1
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
9m 13s
Cost:
$0.146982
Provider:
Together
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.20%
Dialogue to Total Word Ratio
(details)
99.6016%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 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 adjusted the binoculars, the cheap plastic digging into her nose. Through the wet windshield, the house blazed with light. Silhouettes of tiny children bounced inside, unaware of impending doom. "There," Molly whispered. "The loot is next to the piano." Bill squinted, leaning over the console and fogging the glass. He wiped it away with his sleeve. "I see it. A mountain of merchandise. Are we sure they didn't hire a guard?" It was a child's birthday party. The only guard was a retriever named Biscuit. "Biscuit could be a problem. Dogs sense fear." Molly turned her head slowly, staring at him. "You're afraid of a dog named Biscuit?" Bill shifted in his seat, his hand dropping defensively to the wheel. "I fear psychological warfare. What if he barks, and the kids cry? Crying kids ruin timelines." Molly sighed, dropping the binoculars onto the dash. She pulled a blueprint from her pocket, though it was just a crayon drawing. "Focus. We go in the side gate at fourteen hundred hours. That is when the clown arrives." Bill nodded seriously, stroking his chin. "The clown triggers the fire alarm? Or just makes balloon animals until we clear perimeter?" Molly tapped the drawing with a stern finger. "He makes balloon animals. We slip past while he twists a giraffe. You grab loot, I keep lookout, we vanish before cake." Bill frowned, crossing his arms over his jacket. "What about the piƱata? I am not dodging a blindfolded kid with a bat. Kids are ruthless." She held up a small spray bottle, shaking it gently. "Bear mace?" Molly grinned, her teeth flashing in the dim light. "Silly String. If they get aggressive, we blind them with pink foam and run." Bill nodded, making a mental note. "Clever. But what if parents intervene? Moms are vicious. I saw a mom fight another over parking." Molly reached into the glove compartment, pulling out a speaker. "We use the crying baby recording. It clears a room in three seconds. Nobody resists investigating a crying baby." Bill checked his watch, calculating the timeline. "What about the getaway? The driveway is blocked by a bouncy castle. We cannot drive over a bouncy castle." Molly stopped, her mouth open. She looked at the enormous inflatable structure blocking the garage. It was a tactical oversight. "Damn it. I didn't account for tactical inflation. We go on foot. Have skateboards?" Bill popped the trunk, and the van rattled as two rusty skateboards tumbled out into the rain. "Two skateboards, one Silly String can, and a fake mustache. Are we really doing this?" Molly checked her watch, her face set with grim determination. The stakes had never been higher. "We are doing this. For the loot. For the glory. For the Lego Death Star." Bill pulled a fake mustache from his pocket, pressing it firmly over his lip. "Let us go out and make kids cry." They kicked open the van doors and rolled into the rain, professionals on a dastardly mission.
created_at
"2026-04-24T17:20:50.682Z"
4/24/2026, 5:20:50 PM
model
"z-ai/glm-5.1-20260406"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
365
moderation_latency
(null)
generation_time
438469
tokens_prompt
77
tokens_completion
25389
native_tokens_prompt
66
native_tokens_completion
33384
native_tokens_completion_images
(null)
native_tokens_reasoning
24595
native_tokens_cached
0
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
0.146982
router
(null)
provider_responses
0
endpoint_id
"0a3f7e48-cd4f-4fcf-88dd-3ab61bb5e7d1"
id
"ogSwALf-6z2FuN-9f16c7ecf0bc05ab"
is_byok
false
latency
365
model_permaslug
"z-ai/glm-5.1-20260406"
provider_name
"Together"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777051250-bMbrIvlPx7yiCF5qapTA"
session_id
(null)
api_type
"completions"
id
"gen-1777051250-KrNZekiSDGiMpXlrxdHP"
upstream_id
"ogSwALf-6z2FuN-9f16c7ecf0bc05ab"
total_cost
0.146982
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Together"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
99.20%
Dialogue to Total Word Ratio
Ratio: 49.60%, Deviation: 0.40%
neededClean
false
wordsTotal
500
wordsDialogue
248
99.6016%