NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-09-01KX40YDSW758AHFHQQX3S3ZTZ
Run Details
Model:
GPT-5.6 Terra (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 50% dialogue
Duration:
32.6s
Cost:
$0.056940
Provider:
OpenAI
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
100.00%
Dialogue to Total Word Ratio
(details)
100.0000%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 50% 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
Rain polished the suburban cul-de-sac until every parked car resembled a guilty black beetle. Molly crouched behind a hedge, wearing a caterer's apron over her burglar's black sweater. Bill wore a cardboard crown, a fake moustache, and the expression of a man already apologising to prison soup far too early. “Molly, why target a seven-year-old's birthday party instead of bank?” “Banks have guards, cameras, alarms, and disappointingly few inflatable dragons.” “I thought our profession required standards, discretion, and perhaps trousers.” “My standards are high; I rob parties with proper cakes.” “And these presents belong to children, mostly children called Poppy.” “Poppy owns ponies, servants, and an enormous garden telescope too.” Beyond the hedge, Poppy's garden blazed with paper lanterns and weaponised cheerfulness. A plastic pirate ship held juice boxes in its cannons. Near the patio, a magician pulled coloured scarves from his sleeve while parents applauded as though rescuing civilisation. Gift bags glittered beneath a trestle table by the gate. “Rich children deserve character-building losses; mine built character daily beautifully.” “Your childhood involved stealing coal, not character-building birthday larceny exactly.” “Coal never screamed when unwrapped, Bill, unlike this mechanical unicorn.” “That unicorn costs more than mother earned before abandoning me.” “We need an entrance, an exit, and no moral discussion.” “The entrance is caterer van; exit is bouncy castle obviously.” Molly unfolded a hand-drawn plan, which looked less like architecture than a menu attacked by a spider. Bill studied it solemnly, nodding whenever she tapped an arrow. The arrows led from the gate to the presents, then toward a drawing of an airborne castle with wheels and a tiny flag. “A bouncy castle is not transport; it is screaming furniture.” “Precisely, nobody notices screaming furniture crossing the garden fence tonight.” “What about parents, clowns, and the magician with suspicious eyebrows?” “Parents drink prosecco; clowns frighten themselves; magicians respect professional colleagues.” “You have never produced a rabbit, Bill, merely parking tickets.” “Tickets vanish mysteriously, which is essentially magic with municipal paperwork.” Their target was not every present, Molly had explained, only the expensive ones: drones, consoles, jewel-encrusted craft sets, and anything requiring batteries larger than a potato. Bill had privately wondered how one distinguished conscience from inventory management, but Molly's pencil had drawn a tiny skull beside the word ethical twice. “Fine. I will distract everyone by judging the cake competition.” “Excellent; insult frosting, acquire van keys, and avoid the mascot.” “Why avoid him? He is a cheerful, oversized badger surely.” “Last week, mascot arrested me; beneath its head was police.” “That seems unlikely, even for your unusually theatrical criminal career.” “Molly, crime is theatre, except reviews arrive wearing handcuffs afterwards.” A shriek rose from the lawn as the hired badger mascot tripped over a balloon and collapsed into the jelly table. Children cheered. Adults laughed with the relieved hunger of people delighted somebody else had become ridiculous. Molly watched the chaos with professional affection; Bill watched the badger's paws nervously. “Then tonight we steal quietly, swiftly, and with excellent reviews.”
created_at
"2026-07-09T18:06:24.58Z"
7/9/2026, 6:06:24 PM
model
"openai/gpt-5.6-terra-20260709"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
3431
moderation_latency
(null)
generation_time
32539
tokens_prompt
77
tokens_completion
1415
native_tokens_prompt
66
native_tokens_completion
3785
native_tokens_completion_images
0
native_tokens_reasoning
3106
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
0.05694
router
(null)
provider_responses
0
endpoint_id
"3cc89000-ae34-4dde-9c7d-5ce009c8af0b"
id
"resp_028e5f8335b81c56016a4fe320a54c819698487d935a17078a"
is_byok
false
latency
340
model_permaslug
"openai/gpt-5.6-terra-20260709"
provider_name
"OpenAI"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1783620384-ZFbtcD0nLwTfkwtx8mzE"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1783620384-ExS5bfRP0uJtV8CS4pKj"
upstream_id
"resp_028e5f8335b81c56016a4fe320a54c819698487d935a17078a"
total_cost
0.05694
cache_discount
(null)
upstream_inference_cost
0
provider_name
"OpenAI"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
100.00%
Dialogue to Total Word Ratio
Ratio: 50.00%, Deviation: 0.00%
neededClean
false
wordsTotal
508
wordsDialogue
254
100.0000%