NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-18-01KM07DMFSVY4RPMPRJEJ5DVDB
Run Details
Model:
GPT-5.4 Nano (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
10.8s
Cost:
$0.002474
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
50.0000%
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: Bill, whisper. The party is tomorrow, but our courage is tonight. Bill: Tonight? My nerves are doing backflips in my stomach. Molly: Good. The guard dog is imaginary, like my last diet. Bill: Do we wear disguises? Molly: Yes: you’re a clown. I’m a balloon animal. Bill: I can’t juggle. Molly: Neither can the toddlers. That’s why it’s believable. Bill: How do we enter? Molly: Through the side gate. I already crocheted a decoy key from guilt. Bill: Crocheted? Molly: Look, it’s not illegal if it’s artisanal. Bill: And the presents? Molly: We glide to the “From Santa” pile, fill our bags, and bow like we’re disappointed. Bill: Disappointed? Molly: Santa’s gonna ask questions. We’ll say, “We only take the ones with the stickers we like.” Bill: What about the kids? Molly: We distract them with your tragic magic trick. Bill: It’s just me dropping a wand. Molly: Perfect. While they laugh, we steal quietly. Bill: Quietly is my worst skill. Molly: Then be loud in a way that sounds like clapping. Ready? Bill: Ready. If we get caught, I’m blaming the balloons. Molly: Smile, Bill. The best heists end with a napkin tied around your conscience by sunrise.
created_at
"2026-03-18T10:21:20.262Z"
3/18/2026, 10:21:20 AM
model
"openai/gpt-5.4-nano-20260317"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
9639
moderation_latency
(null)
generation_time
10650
tokens_prompt
77
tokens_completion
330
native_tokens_prompt
66
native_tokens_completion
1969
native_tokens_completion_images
0
native_tokens_reasoning
1634
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
0.00247445
router
(null)
provider_responses
0
endpoint_id
"0c835f2e-c18d-4e8c-b245-e1e3bd08b97f"
id
"resp_0d8bfbe75a512f740169ba7ca15cec819794244ea1607a807a"
is_byok
false
latency
1238
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-1773829280-UGeAOM3gb2Q40pNm4E96"
upstream_id
"resp_0d8bfbe75a512f740169ba7ca15cec819794244ea1607a807a"
total_cost
0.00247445
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
200
0.00%
Dialogue to Total Word Ratio
Ratio: 6.00%, Deviation: 84.00%
neededClean
false
wordsTotal
200
wordsDialogue
12
50.0000%