NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ07VKNVA70CW73HX0AQ95HV
Run Details
Model:
Z.AI GLM 5.1
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
2m 9s
Cost:
$0.054608
Provider:
AtlasCloud
Evaluation summary
Score
Evaluator
Details Link
1.04%
Matches word count
(details)
85.34%
Dialogue to Total Word Ratio
(details)
43.1903%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 tapped the steering wheel, her eyes locked on the house across the street. The front yard had been transformed into a chaotic sprawl of primary colors. A massive inflatable castle heaved in the wind, its plastic walls squeaking as a horde of small children ricocheted inside. Two ponies stood tethered to the mailbox, looking profoundly depressed. The shrieks of sugared-up kids carried faintly through the van's cracked window. "The target is mobile. She's putting a large, glittering box right on the prize table." "You think it's a pony?" "Nobody boxes a pony, Bill. It's probably a bicycle. Or an elaborate Lego set." "Lego sets are where the money is. The secondary market for minifigures is insane right now." Bill noted, unwrapping a granola bar. "I know. That's why we're here. The parents are in the kitchen. The kids are in the bouncy castle. The gift table is by the sliding glass doors." "What about the dad? I saw a dad out there." "He's manning the grill. Once the hot dogs are ready, he'll be completely distracted." "What if he burns the hot dogs? Then he comes inside to complain." "Then we offer him a fresh hot dog, Bill." "We're bringing hot dogs?" "No! We are bringing duffel bags. Remember the plan? In through the side gate, grab the gifts, out through the side gate. We do not speak to parents. We do not pet the ponies. We do not eat the cake." "Even if it's red velvet?" "Even if it's red velvet." "You're a monster, Molly." Bill nodded sagely, brushing crumbs off his black turtleneck. He considered himself a connoisseur of the illicit plastic brick trade. "I'm a professional. Now, what's our biggest obstacle?" "The clown." "Exactly. Clowns are unpredictable. They make balloon swords. They do magic tricks. They draw attention. When the clown starts his big finale, that's our window." "What's his big finale?" "He pulls an endless string of flags out of his pants." Molly rubbed her temples, feeling a migraine forming behind her left eye. "And that distracts the kids?" "It terrifies them. The screaming will draw all the adults. That's when we move. Grab the biggest boxes first. Leave the envelopes. We don't have time to check for gift cards." "I can check for gift cards while you load the bags. I'm very fast with envelopes." "You are not checking envelopes. Do you know how long it takes to check an envelope?" "Three seconds. Four if it's sealed tight." "Bill, we have a forty-second window. We grab boxes, we go. No envelopes. No cake. No clowns. Just grab and dash." "Fine. But if a kid throws a punch, I'm defending myself." "A kid is not going to punch you." "You don't know. These kids today are feral. I saw one bite a clown last week." Outside, a rusted car pulled up. A clown emerged, his massive shoes stomping onto the driveway. "That clown deserved it. Now grab the bags. The clown is reaching for his pants." Bill stuffed the rest of his granola bar into his mouth, grabbed a heavy duffel bag from the backseat, and quietly popped the passenger door.
created_at
"2026-04-24T17:16:03.669Z"
4/24/2026, 5:16:03 PM
model
"z-ai/glm-5.1-20260406"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
2636
moderation_latency
(null)
generation_time
128620
tokens_prompt
77
tokens_completion
11240
native_tokens_prompt
66
native_tokens_completion
12390
native_tokens_completion_images
(null)
native_tokens_reasoning
11684
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
0.05460774
router
(null)
provider_responses
0
endpoint_id
"7e919c5e-6c06-4182-adaf-2910d5d4b5a8"
id
"0591e9d097b84080b96d2ffcdc187bb9"
is_byok
false
latency
2631
model_permaslug
"z-ai/glm-5.1-20260406"
provider_name
"AtlasCloud"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777050963-KC49454NZIDe0rYDw5Yn"
session_id
(null)
api_type
"completions"
id
"gen-1777050963-EaIptMgpLtgy6oTANk1x"
upstream_id
"0591e9d097b84080b96d2ffcdc187bb9"
total_cost
0.05460774
cache_discount
(null)
upstream_inference_cost
0
provider_name
"AtlasCloud"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
1.04%
Matches word count
n/a
neededClean
false
words
526
85.34%
Dialogue to Total Word Ratio
Ratio: 71.78%, Deviation: 1.78%
neededClean
false
wordsTotal
528
wordsDialogue
379
43.1903%