NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ0NZM8373M6KDGA4KEBDZX8
Run Details
Model:
GPT-5.5 (Reasoning, Low)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
19.8s
Cost:
$0.039810
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
20.19%
Matches word count
(details)
99.99%
Dialogue to Total Word Ratio
(details)
60.0900%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 spread the birthday invitation across the dashboard like a treasure map, flattening cupcake crumbs with one black-gloved finger. Bill squinted through the rain at the community hall, where balloons bobbed in the windows and a papier-mâché dragon guarded the door with more dignity than most security guards. “We go in as magicians,” Molly said. “You can be the rabbit.” Bill patted his stomach. “I’m allergic to applause.” Inside, shrieks of musical chairs shook the glass. Parents milled about, dazed and sticky, perfect cover. On the curb, a pony in a party hat stared at their van as if memorizing the license plate. Molly had calculated everything: three minutes during cake, two sacks per trip, exit through the kitchen, apology note if necessary. Bill had contributed a map drawn on a napkin, mostly arrows and a surprisingly detailed sausage roll. Then the hall doors burst open. A dozen children charged out, each clutching a wrapped present, pursued by a clown shouting about fairness. The gifts were being redistributed for a game. Molly folded the invitation. Bill started the engine. “New plan,” she said. Bill adjusted mirrors and cleared confetti from the gearshift, trying not to look relieved. “Finally, adult supervision.” “Hardly,” Molly said. They followed the pony, who looked like management material. Behind them, the dragon sagged in soggy judgment.
created_at
"2026-04-24T21:22:55.382Z"
4/24/2026, 9:22:55 PM
model
"openai/gpt-5.5-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
17188
moderation_latency
(null)
generation_time
18909
tokens_prompt
77
tokens_completion
776
native_tokens_prompt
66
native_tokens_completion
1316
native_tokens_completion_images
0
native_tokens_reasoning
1018
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"
usage
0.03981
router
(null)
provider_responses
0
endpoint_id
"58e5b336-423e-430b-a2ab-8bc353f0c51b"
id
"resp_0e90d2ca1fddb0b70169ebdf2f77a081978ea07b77def60039"
is_byok
false
latency
235
model_permaslug
"openai/gpt-5.5-20260423"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777065775-v9smtI81AIC5odVAgEhF"
session_id
(null)
api_type
"completions"
id
"gen-1777065775-Dxyyh1lGaxXWAgX4Gz0v"
upstream_id
"resp_0e90d2ca1fddb0b70169ebdf2f77a081978ea07b77def60039"
total_cost
0.03981
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
20.19%
Matches word count
n/a
neededClean
false
words
220
99.99%
Dialogue to Total Word Ratio
Ratio: 9.01%, Deviation: 0.99%
neededClean
false
wordsTotal
222
wordsDialogue
20
60.0900%