NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36NKDSG6C285WCW1RED40K5
Run Details
Model:
GPT-6 Luna
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
9.9s
Cost:
$0.000394
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
21.85%
Dialogue to Total Word Ratio
(details)
10.9250%
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 unrolled the floor plan across the bonnet of Bill’s car. A cartoon rabbit smiled up at them from the page. “Why does the venue have a rabbit logo?” Bill asked. “Because it’s a children’s party.” “No, I mean why is the rabbit wearing a waistcoat? He looks like he knows where the presents are.” Molly pinned the paper with a thermos. “The presents are in the community hall, beside the stage. We go in through the kitchen, cross the corridor, and leave through the fire exit.” “Stealing birthday presents from children,” Bill said. “At last, a crime with no moral complications whatsoever.” “Think of it as temporary redistribution.” “To whom?” “Us.” Bill studied the plan. “What’s the code word?” “Cupcake.” “What if someone offers us one?” “Then say no.” “What if it’s chocolate?” “Bill.” “I’m checking the security protocol.” Inside the hall, a burst of laughter rose above the music. Molly peered through the kitchen window. Bright paper streamers drooped from the ceiling. A dozen children raced in circles while a tired parent held a tray of juice boxes. “Kitchen’s clear,” she whispered. “Except for the woman with the tray.” “She’s going to the tables.” “And the enormous man dressed as a rabbit?” Molly lowered the binoculars. “That’s the entertainer.” “He has a waistcoat.” “I know.” “He’s guarding the presents.” “He’s leading a conga line.” The rabbit bounced past the window, followed by children waving paper crowns. Bill watched them go. “Could we perhaps steal the rabbit?” “No.” “Fine. But I want it noted that he’s the most suspicious person here.” They slipped through the kitchen door. Molly wore a flowered dress and a plastic tiara; Bill carried a tray of cupcakes. “Why am I holding these?” he murmured. “Because guests carry things.” “Guests carry gifts. These are cupcakes.” “Then look generous.” “I look like I’ve been assigned a task.” “That’s what generosity looks like.” They entered the corridor. At its end stood a table piled with brightly wrapped presents, each labeled in careful marker. Bill slowed. “Those are smaller than I expected.” “They’re for children.” “I expected more gold.” Molly reached for a silver parcel. “Take the whole pile.” A little girl appeared behind them, clutching a balloon. “Are you the party helpers?” Bill froze. “We’re the… presents department.” “That’s not a department,” said Molly. “It is now.” The girl pointed at his tray. “Can I have a cupcake?” Bill looked at the cupcake, then at Molly. “Protocol,” she hissed. He handed it over. “Cupcake.” The girl beamed and ran back to the party. Molly exhaled. “We’re leaving.” “But the presents—” “Not ours.” “You said temporary redistribution.” “I’ve reconsidered the policy.” They turned toward the exit. The rabbit entertainer blocked it, holding a paper plate. “Leaving already?” he asked. Bill raised the empty tray. “We’ve distributed the cupcakes.” “Excellent,” said the rabbit. “You’re the best volunteers we’ve had all afternoon.” Outside, Molly glanced back through the window. The children had gathered around the presents, shrieking as they tore the paper. Bill sighed. “I suppose we did help.” “We committed no crime.” “We entered under false pretenses.” “We wore costumes.” “You stole a cupcake.” “I donated it.” Bill considered this. “Next week, there’s a retirement party.” Molly started the car. “We’re not stealing from retirees.” “Of course not. We’ll be the entertainment.”
created_at
"2026-09-23T08:20:04.023Z"
9/23/2026, 8:20:04 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1875
moderation_latency
(null)
generation_time
9849
tokens_prompt
75
tokens_completion
1019
native_tokens_prompt
66
native_tokens_completion
774
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
0.0003936
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.20.0; linux; x64))"
http_referer
(null)
request_id
"req-1790151604-BuGKdXg77V3FPMjAIRYY"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790151604-Il2i37bVuc7XEftf2nDq"
upstream_id
"resp_03f921a3e735ebed016ab38bb41f5c87d194e879c884e46055"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_03f921a3e735ebed016ab38bb41f5c87d194e879c884e46055"
is_byok
false
latency
300
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0003936
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
553
21.85%
Dialogue to Total Word Ratio
Ratio: 55.52%, Deviation: 5.52%
neededClean
false
wordsTotal
553
wordsDialogue
307
10.9250%