NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-24-01KQ0NCPVQS256948XWGQWQ46T
Run Details
Model:
GPT-5.5 (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
23.2s
Cost:
$0.050130
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
1.61%
Dialogue to Total Word Ratio
(details)
50.8025%
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 crinkled party invitation across the café table like a treasure map, ignoring the frosting smear shaped suspiciously like Portugal. Bill leaned in, wearing a paper crown he'd "liberated" from the display stand, and tried to look tactical instead of adorable. The target was the Peppercorn twins' seventh birthday, a balloon-choked extravaganza at the church hall, rumored to contain twelve wrapped dinosaurs, three glitter laboratories, and one suspiciously heavy box labeled Educational Socks. Molly tapped the hall's sketch with a teaspoon. "We enter as magicians, grin, and vanish gifts before cake." Bill frowned at a doodled clown. "Fine, but no crawling through anything labeled Fun, ever again." Their plan depended on chaos: sugared children stampeding after a piñata, parents debating allergies, and Uncle Graham performing accordion versions of pop songs until mercy became a public issue. Molly would distract the bouncy-castle attendant with fake coupons; Bill would wheel in a laundry cart marked Lost Mittens. Neither mentioned what they would do if a child cried. That was the flaw in every villainous enterprise: tiny faces, enormous eyes, and moral collapse. Still, Molly packed confetti smoke pellets anyway. Bill added sandwiches, because heists required snacks and plausible crumbs afterward.
created_at
"2026-04-24T21:12:35.455Z"
4/24/2026, 9:12:35 PM
model
"openai/gpt-5.5-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
21379
moderation_latency
(null)
generation_time
22939
tokens_prompt
77
tokens_completion
734
native_tokens_prompt
66
native_tokens_completion
1660
native_tokens_completion_images
0
native_tokens_reasoning
1389
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
0.05013
router
(null)
provider_responses
0
endpoint_id
"58e5b336-423e-430b-a2ab-8bc353f0c51b"
id
"resp_08346fe062a755e10169ebdcc398848194bacc90bbad3cec19"
is_byok
false
latency
488
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-1777065155-1qF54aQ2gmwr83AMopAg"
session_id
(null)
api_type
"completions"
id
"gen-1777065155-tgzNaJU37HB3cXkGcKpV"
upstream_id
"resp_08346fe062a755e10169ebdcc398848194bacc90bbad3cec19"
total_cost
0.05013
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
200
1.61%
Dialogue to Total Word Ratio
Ratio: 24.26%, Deviation: 14.26%
neededClean
false
wordsTotal
202
wordsDialogue
49
50.8025%