NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-18-01KM07CKM769PVDAZADF33E9Q6
Run Details
Model:
GPT-5.4 Nano (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
13.4s
Cost:
$0.003783
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.99%
Dialogue to Total Word Ratio
(details)
99.9935%
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
"Molly: Okay, Bill, the toddler party is tonight." "Bill: Perfect. I brought silent roller skates, like the legends." "Molly: Legends? We’re stealing presents, not escaping a museum." "Bill: Shh. The clown is doing balloon animals. We must look like confused parents." "Molly: I can do confused; I practiced in the grocery store aisle with juice boxes." "Bill: Great. I’ll be “Daddy Bill,” wearing my heroic mustache." "Molly: No mustache. That’s a beacon." "Bill: Fine. I’ll hide it under this hat shaped like a question mark." "Molly: Why is the plan always weirder?" "Bill: Because security is psychological. Kids only fear nonsense." "Molly: So we’ll grab all presents before cake time." "Bill: Yes, but shiny paper only. The rest look like homework." "Molly: Don’t insult the toys." "Bill: I never do. I insult wrapping." "Molly: Door code?" "Bill: Invitation said, “Follow the glitter.”" "Molly: Glitter leads to the living room. Then we bag everything and run." "Bill: If a child asks questions?" "Molly: We say, “Surprise delivery!”" "Bill: Works every time. Now listen—someone’s coming." "Molly: Breathe, Bill. Act normal." "Molly: If the presents are heavy, kick the box, never the kid; Bill, agree please?" "Bill: Normal is impossible. Stealing is forever."
created_at
"2026-03-18T10:20:46.61Z"
3/18/2026, 10:20:46 AM
model
"openai/gpt-5.4-nano-20260317"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
12459
moderation_latency
(null)
generation_time
13284
tokens_prompt
77
tokens_completion
328
native_tokens_prompt
66
native_tokens_completion
3016
native_tokens_completion_images
0
native_tokens_reasoning
2679
native_tokens_cached
0
num_media_prompt
(null)
num_input_audio_prompt
(null)
num_media_completion
0
num_search_results
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"completed"
usage
0.0037832
router
(null)
provider_responses
0
endpoint_id
"0c835f2e-c18d-4e8c-b245-e1e3bd08b97f"
id
"resp_0931b6b753beb44a0169ba7c7eac1c8190859dd518da474614"
is_byok
false
latency
235
model_permaslug
"openai/gpt-5.4-nano-20260317"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1773829246-NaTsPrDxIeGE7aIjeASR"
upstream_id
"resp_0931b6b753beb44a0169ba7c7eac1c8190859dd518da474614"
total_cost
0.0037832
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
201
99.99%
Dialogue to Total Word Ratio
Ratio: 91.04%, Deviation: 1.04%
neededClean
false
wordsTotal
201
wordsDialogue
183
99.9935%