NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-02-01M1HGBHRMFYXMDYKDEX9VHYXD
Run Details
Model:
Qwen 3.8 Flash (Reasoning)
Test:
Dialogue tags
Scenario:
Write 500 words with 30% dialogue
Duration:
4m 60s
Cost:
$0.011126
Provider:
Alibaba
Evaluation summary
Score
Evaluator
Details Link
100.00%
Matches word count
(details)
98.78%
Dialogue to Total Word Ratio
(details)
99.3911%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 30% 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 a glittery party invitation across the van dashboard like a war map, while Bill tried to fold a napkin into a stealth mask and failed immediately. The invitation announced a unicorn-themed birthday at a suburban house, complete with cake, piñata, and one hundred presents. "Bill, this is the heist we were born for. All wrapped, unsupervised, probably near a bathroom." "Can we call it a rescue operation?" "No." "Heist." The cartoon pony stared back at them, bright with innocence and absolutely no security expertise. A clown waited outside, again. "The plan is simple. We slip in during the sing-along, gather every present by color and size, then vanish before anyone notices the cake has become a crime scene." Bill nodded with the gravity of a man who had once robbed a bakery by mistake. Outside, balloons bumped the windows cheerfully, too close. "What if a kid sees us?" "Then you distract them." "With what?" "Your face." They watched the patio from behind hedges. A toddler waved at a duck. A teenager looked bored enough to be stolen. The presents sat beneath a weighted tablecloth. "I have a trustworthy face." "It’s a hostage negotiation nightmare." A clown argued with helium near the castle, which Molly noted as either flaw or accomplice in training. "We’ll need carts." "Use the gift table." "That’s suspicious." "Bill, we are stealing forty presents." "Forty-three." "Don’t correct me mid-crime." At three fourteen, music changed. Molly checked her watch, then her reflection. She wore black gloves, a mask, and a pinned aggressive birthday card. "I made this backup gift in case we get caught." "You’re not a hostage negotiator. You’re a hostage with anxiety." Bill produced a gift wrapped in newspaper. It had one ribbon, two typos, and the nervous energy of a man who had rehearsed being nice in public places before this. "Great. On three?" "On forty-three." "I love that number." They approached the table like two raccoons who had discovered capitalism, each carrying a tote bag and a regret about children’s music so loud. "We need a calling card. It should say thank you for nothing." "Mystery." "Humor." Molly counted gifts as if they were diamonds. Bill stacked boxes by color, size, and suspected weight. They moved like weather, whispering insults at ribbon until the patio began to clear early. Only one enormous box remained, tied with a bow so large it seemed designed for a small horse. Molly stared. Bill poked it again. "No." Bill lifted the flap. A false bottom gave way beneath his napkin mask. Molly caught his collar, then lost it, then caught it again with one hand while holding a tote bag of balloons. "Did we win?" The box tilted. The bow wobbled. Molly realized with horror that it had not hidden the present. It had hidden them too. Then the cardboard folded upward like a guilty magic trick, swallowing the napkin mask, the gloves, and both professional criminals without asking permission again. "We are in the box."
created_at
"2026-09-02T16:48:38.174Z"
9/2/2026, 4:48:38 PM
model
"qwen/qwen3.8-flash-20260826"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
731
moderation_latency
(null)
generation_time
299487
tokens_prompt
75
tokens_completion
17998
native_tokens_prompt
127
native_tokens_completion
23632
native_tokens_completion_images
(null)
native_tokens_reasoning
22944
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
0.01112609
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788367718-X7ouAiChb48V1cV7UP7n"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788367718-4ahGTIirAsopkHBagzGT"
upstream_id
"chatcmpl-9f112f16-4582-9eb6-925f-52d192014fac"
provider_responses
0
endpoint_id
"84b1e4a7-5aed-4464-818f-1994f0b4ee18"
id
"chatcmpl-9f112f16-4582-9eb6-925f-52d192014fac"
is_byok
false
latency
731
model_permaslug
"qwen/qwen3.8-flash-20260826"
provider_name
"Alibaba"
status
200
total_cost
0.01112609
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
98.78%
Dialogue to Total Word Ratio
Ratio: 30.50%, Deviation: 0.50%
neededClean
false
wordsTotal
505
wordsDialogue
154
99.3911%