NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HCWGW98M2KK0N2XXMF781T
Run Details
Model:
Qwen 3.8 Max (Reasoning, XHigh)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
5m 27s
Cost:
$0.093620
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
99.93%
Dialogue to Total Word Ratio
(details)
99.9643%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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
The inflatable castle wheezed in the yard like a pink marshmallow with secrets. Molly crouched behind a hedge shaped like a rabbit that had clearly given up. Bill held a birthday hat on his knee like a hostage. “So,” Molly said, “the presents are in the gazebo.” “Not gazebo,” Bill said. “There is a difference. Gazebos are for people who trust birds.” “Fine. Pavilion. Guarded by one woman named Brenda and forty children whose hands are sticky for no legal reason.” “Brenda is the real threat. She has a clipboard.” “Clipboards can be defeated.” “Not this one. It has glitter. It has a seating chart. It has a schedule that says cake at four, which means presents at four-twenty. We move at four-eighteen.” “Why not during cake?” “Because cake makes them emotional. Emotional children defend wrapping paper with surprising ferocity.” Molly adjusted her binoculars, which were actually two toilet rolls painted silver. “What’s the entry?” she asked. “We pose as entertainment.” “We are not entertainers.” “I can juggle.” “You cannot.” “I can juggle softly.” “Bill.” “Fine. We pose as the balloon person.” “There is already a balloon person.” “Then we pose as the balloon person’s accountant.” “Nobody hires an accountant for balloons.” “That is why it is brilliant. Nobody questions a balloon accountant. They assume the math is depressing and let us pass.” A child screamed somewhere beyond the hedge, followed by the sound of a plastic trumpet being used as a weapon. “We need disguises,” Molly said. “I brought a moustache.” “It is a caterpillar.” “It is a very convincing caterpillar.” “Why do you have it?” “I found it near the pet shop job.” “Put it away. We’ll use the piñata.” “Steal the piñata?” “No. Replace it.” “With what?” “Ourselves.” “Molly, we cannot fit in a piñata.” “We can fit in a large piñata.” “This is a unicorn piñata.” “Then we shall be a very ambitious unicorn.” Bill looked at the pavilion, where a table of wrapped boxes gleamed under crepe paper. “I love this season,” he said. “It is July.” “I mean birthday season.” “Focus. You distract Brenda.” “With what?” “Compliments.” “I am not built for compliments.” “You once told a security guard his flashlight had excellent posture.” “That was true.” “Then be true again.” He straightened his tie, which was actually a ribbon stolen from a bakery. “And if she resists?” he asked. “Deploy the cupcakes.” “Those were for morale.” “All cupcakes are for morale, Bill. Some are also tactical.” Molly peeked over the hedge as a balloon squeaked against a stack of presents. “Bill, if we get caught, we are surprise clowns.” “I hate clowns.” “You are one.” “I prefer respectful crime.” “Crime is not respectful.” “It can wave.” “It can bow.” The piñata swayed above the lawn, full of candy and poor decisions. “On three,” Molly said. “Why three?” “One is too eager. Two is suspicious. Three sounds official.” “Fine. One, two—” “Three.” They charged toward the pavilion, armed with ribbon, crumbs, and absolute confidence.
created_at
"2026-09-02T15:48:00.017Z"
9/2/2026, 3:48:00 PM
model
"qwen/qwen3.8-max-20260803"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1170
moderation_latency
(null)
generation_time
326960
tokens_prompt
75
tokens_completion
11413
native_tokens_prompt
127
native_tokens_completion
15561
native_tokens_completion_images
(null)
native_tokens_reasoning
14782
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
"stop"
service_tier
(null)
usage
0.09362
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788364080-3fhBqfrvAyPukjVY9wKg"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788364080-0kjMfFkAmq5jqFm3DJdI"
upstream_id
"chatcmpl-1d0b63b5-7316-95c5-a8d1-c9d68ea5cfba"
provider_responses
0
endpoint_id
"6332e9fd-e0c4-4cae-8c17-d25bc774864a"
id
"chatcmpl-1d0b63b5-7316-95c5-a8d1-c9d68ea5cfba"
is_byok
false
latency
1170
model_permaslug
"qwen/qwen3.8-max-20260803"
provider_name
"Alibaba"
status
200
total_cost
0.09362
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Alibaba"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Matches word count
n/a
neededClean
false
words
500
99.93%
Dialogue to Total Word Ratio
Ratio: 70.12%, Deviation: 0.12%
neededClean
false
wordsTotal
502
wordsDialogue
352
99.9643%