NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-23-01M36R6QGFV3B90QXZF3DEYT0B
Run Details
Model:
GPT-6 Luna (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
1m 55s
Cost:
$0.005543
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.29%
Dialogue to Total Word Ratio
(details)
99.6441%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 studied the community-hall window through a pair of opera glasses, though the hall was only across the street and the glasses were on backward. Inside, a paper banner announced NED’S SIXTH BIRTHDAY. Beneath it stood a mountain of brightly wrapped presents, guarded by one sleepy balloon shaped like a dinosaur. “Suspicious disguise,” Bill whispered. “You look like a duchess spying on a window.” “I’m blending in,” Molly said. “Duchesses adore balloons.” Bill wore a paper crown from a bin marked TAKE ONE; Molly objected on principle. “Simple,” Bill said. “We enter, collect the presents, leave.” “That’s a sentence, not a plan,” Molly said. “Then we enter politely.” “Thieves don’t.” They waited for the party’s song to end. A chorus of children shouted the birthday wish; then the door swung open and a boy emerged carrying a cake plate. Molly and Bill simultaneously attempted to look like ordinary adults who had always been standing beside a shrub. “Are you here for Ned’s party?” he asked. “Absolutely,” said Bill. “We’re present inspectors.” “Is that real?” “In exclusive neighborhoods,” Molly said. The boy studied them. “You look like people who lost a bet.” “That’s on our résumé,” said Molly. He invited them in before either could invent a credential. The room smelled of frosting, orange juice, and the particular panic of adults who had just noticed two unfamiliar grown-ups near the gifts. Bill whispered that this was going wonderfully. Molly whispered that he had confused hospitality with security. “Now,” Bill murmured, “we take the presents.” “Where?” “Into the night. Quickly. Silently.” “It’s four in the afternoon.” “Late afternoon.” “Grandmother’s watching.” “She looks kind.” “She has a serving spoon like a gavel.” Molly’s gaze settled on the presents, then on the children’s handmade name tags. Ned had drawn a crooked star on each one. She pictured the disappointment of finding an empty pile and felt her confidence leak out like air from Bill’s crown. The crown sagged over one eyebrow. “Maybe we should leave,” Molly said. “With no loot?” “With dignity.” “I left mine in the bin.” “Then take your crown.” “It’s cardboard.” “Exactly. Let’s go.” Before Bill could answer, the grandmother approached and pressed a paper plate into his hands. “Could you two help serve cake?” she asked. Bill glanced at Molly. Molly glanced at the gifts. Then she took the plate from him and began cutting generous slices. Ned peered up at Molly. “Can you inspect the cake?” “Certainly. Our findings are serious: frosting, sprinkles, and a suspicious number of candles.” “Potentially dangerous?” Bill asked. “Dangerous to trousers, though,” said Molly. Their heist dissolved into frosting, birthday candles, and a game of musical chairs. Somewhere, a kazoo wheezed the national anthem of poor decisions, bravely. At the end, Ned handed each thief a party bag. Bill examined his carefully: a sticker, a toy whistle, and one tiny chocolate. Molly pocketed it with a wink. He nodded, solemnly, as the dinosaur balloon drifted above them like an accomplice who had turned state’s witness.
created_at
"2026-09-23T09:05:33.723Z"
9/23/2026, 9:05:33 AM
model
"openai/gpt-6-luna-20260922"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
16774
moderation_latency
(null)
generation_time
114445
tokens_prompt
75
tokens_completion
1852
native_tokens_prompt
66
native_tokens_completion
11073
native_tokens_completion_images
0
native_tokens_reasoning
10382
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
0.0055431
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1790154333-OKSCvKyeDESuLjNBO0Mb"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1790154333-WnB8b79igq0Ptl0YAhkY"
upstream_id
"resp_01ec671314d42ac8016ab3965dd2c487d1a67047cea0d5d78c"
provider_responses
0
endpoint_id
"05e94e02-b9c2-4bbb-ba55-4082ee9ad687"
id
"resp_01ec671314d42ac8016ab3965dd2c487d1a67047cea0d5d78c"
is_byok
false
latency
909
model_permaslug
"openai/gpt-6-luna-20260922"
provider_name
"OpenAI"
status
200
total_cost
0.0055431
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
100.00%
Matches word count
n/a
neededClean
false
words
501
99.29%
Dialogue to Total Word Ratio
Ratio: 29.62%, Deviation: 0.38%
neededClean
false
wordsTotal
503
wordsDialogue
149
99.6441%