NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NT6A985EM9NA4RT19B2KX
Run Details
Model:
GPT-6 Luna
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
4.8s
Cost:
$0.000184
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
99.83%
Dialogue to Total Word Ratio
(details)
49.9173%
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
“I’ve studied the party,” Molly whispered. “There are twelve children, one magician, and a piñata shaped like a suspiciously well-armed unicorn.” “Security?” “Two parents, three balloons, and a golden retriever with the moral authority of a judge.” “Then we go in disguised as entertainers.” “You can’t juggle.” “I can panic rhythmically.” “Fine. I’ll be the magician. You’re my assistant.” “Why am I always the assistant?” “Because last time you wore a cape, you got trapped in a revolving door.” “That cape was poorly designed.” “The presents are under the table. We grab them during ‘Happy Birthday.’” “While everyone sings?” “Exactly. Nobody notices anything during the birthday song. It’s forty seconds of social hypnosis.” “And the dog?” “Offer it a cupcake.” “Dogs shouldn’t eat cupcakes.” “Then offer it a lecture about responsible snacking.” Molly peered through the window. “Wait. The birthday child is opening a gift now.” “Already?” “It’s a puppy.” Bill lowered the sack. “We’re not stealing presents from a child who just got a puppy.” “We’re thieves, Bill, not monsters.” “Good. We’ll leave a gift.” “What?” “Our getaway map.” “That’s not a gift.” “It’s handmade.” “Put the sack away. We’ll steal the piñata instead.” “It’s full of candy.” “Exactly. A clean score.” The unicorn burst through the window. Bill stared. “You said piñatas were harmless.” “I said suspiciously armed.” “Run!” “Not with the presents!” “Then run with the candy!” “And the dog?” “Let the judge decide!”
created_at
"2026-09-23T08:23:45.743Z"
9/23/2026, 8:23:45 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1411
moderation_latency
(null)
generation_time
4771
tokens_prompt
75
tokens_completion
471
native_tokens_prompt
66
native_tokens_completion
355
native_tokens_completion_images
0
native_tokens_reasoning
0
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
"completed"
service_tier
"default"
usage
0.0001841
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151825-mUv7gaaTU7ESVnbFkAlC"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151825-WKxlrxwyyCByp3xBL9dt"
upstream_id
"resp_0028894ff591220c016ab38c91d4f487d181b7752123581c47"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_0028894ff591220c016ab38c91d4f487d181b7752123581c47"
is_byok
false
latency
361
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0001841
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
237
99.83%
Dialogue to Total Word Ratio
Ratio: 92.02%, Deviation: 2.02%
neededClean
false
wordsTotal
238
wordsDialogue
219
49.9173%