NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-15-01KKS2Q47T4JX2ZJMEKPV56K6V
Run Details
Model:
Z.AI GLM 5 Turbo
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
8.8s
Cost:
$0.002421
Provider:
Z.AI
Evaluation summary
Score
Evaluator
Details Link
2.01%
Matches word count
(details)
1.20%
Dialogue to Total Word Ratio
(details)
1.6083%
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 smoothed the blueprint across the diner table, her eyes darting toward the waitress before leaning in close. "Right, so the target is a seven-year-old's birthday party at the Henderson residence. Saturday, two o'clock." Bill set down his coffee mug with a skeptical squint. "You want us to rob a children's party." "I want us to liberate high-end presents from a children's party. There's a difference." "Is there?" "Absolutely. We're not taking the cake, Bill. We have standards." Bill gestured at the paper with his fork. "And what exactly are we looking at here? What seven-year-old has anything worth our time?" Molly grinned the kind of grin that had gotten them both into trouble more times than either could count. "Little Timothy Henderson. His mother is Linda Henderson. Linda Henderson who married Richard Henderson. Richard Henderson who owns half the commercial real estate in this city and apparently does not understand appropriate gifting for a child." "How inappropriate are we talking?" "The wish list on the invitation included a PlayStation Five, a drone that retails at eight hundred dollars, and what I believe is a junior gold Rolex." Bill whistled low. "A gold watch for a seven-year-old." "Solid gold, Bill. I checked." "And you think these are just sitting on a present table?" "That is exactly where unwrapped presents sit at children's parties. I have done reconnaissance. There is a bounce house in the backyard, a clown who looks three days from retirement, and a present table inside the sliding glass doors with no line of sight from the parents." "You went to the house." "I went to the house." "Posing as what?" "A concerned neighbor asking about parking restrictions for the party. Linda gave me a full tour. The woman loves to talk about her catering choices." Bill rubbed his temples. "So our extraction point is through the sliding glass doors. We walk in during the chaos of present opening, grab the good stuff, and walk out looking like relatives who forgot to sign a card." "Exactly." "What about the clown?" "What about him?" "Clowns see everything, Molly. It's literally their job to watch people." Molly waved her hand dismissively. "I have a plan for the clown." "Please tell me it does not involve the tear gas." "It does not involve the tear gas." "Then what?" "I have recruited a four-year-old named Destiny from the neighborhood. She will be positioned near the bounce house with instructions to request balloon animals continuously for the duration of the operation." Bill stared at her for a long moment. "You've thought of everything, haven't you?" "I have thought of almost everything. I still need you to confirm you can fit into the inflatable dinosaur costume." Bill looked down at his coffee. "The dinosaur costume." "The dinosaur costume. You'll be the distraction if anything goes sideways. Nobody suspects a dinosaur." "Nobody suspects a dinosaur at a children's party?" "Exactly. You'll blend right in." Bill picked up his mug again, drained the last of his coffee, and sighed with the weight of a man who had made peace with ridiculous things long ago. "When do I pick up the costume?"
created_at
"2026-03-15T15:44:27.392Z"
3/15/2026, 3:44:27 PM
model
"z-ai/glm-5-turbo-20260315"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1545
moderation_latency
(null)
generation_time
8744
tokens_prompt
77
tokens_completion
939
native_tokens_prompt
66
native_tokens_completion
742
native_tokens_completion_images
(null)
native_tokens_reasoning
81
native_tokens_cached
22
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.002420864
router
(null)
provider_responses
0
endpoint_id
"9ce3b132-3a17-4fea-8092-0e8e519ed32b"
id
"20260315234428e840065d43244baf"
is_byok
false
latency
1545
model_permaslug
"z-ai/glm-5-turbo-20260315"
provider_name
"Z.AI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1773589467-2Lvlc2YzHxCUp63wYHI5"
upstream_id
"20260315234428e840065d43244baf"
total_cost
0.002420864
cache_discount
0.00002112
upstream_inference_cost
0
provider_name
"Z.AI"
Evaluation details
Result
Evaluator
Details
Meta Data
2.01%
Matches word count
n/a
neededClean
false
words
525
1.20%
Dialogue to Total Word Ratio
Ratio: 79.40%, Deviation: 9.40%
neededClean
false
wordsTotal
534
wordsDialogue
424
1.6083%