NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-25-01KMJ818N424XVJ7E6D84S8VV8
Run Details
Model:
Qwen3 235B A22B Instruct 2507
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
3.7s
Cost:
$0.000038
Provider:
WandB
Evaluation summary
Score
Evaluator
Details Link
75.16%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
37.5778%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 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
Molly adjusted her ski mask in the rearview mirror. “You sure this clown’s birthday gig is the one?” Bill checked the map on his phone. “Double-Dippy Dino’s Playland. Pink balloon arch, bounce castle, clown with a suspiciously deep voice. That’s the spot.” “That voice *is* suspicious,” Molly muttered. “Could be an inside man.” “They always cry when we take the piñata. Weak stomachs.” “You bring the loot bags?” Bill patted his backpack. “Three XL trash bags. One for toys, one for candy, one for emotional devastation.” “The kids won’t even know we were there.” “Oh, they’ll know. They’ll wake up tomorrow missing their talking unicorn and half a bag of gummy worms. Trauma bonds are real.” “You distract at the dunk tank. I’ll sweep the gift table.” “Why *me* at the dunk tank?” “Because last time you tried to blend in at a preschool, you taught them poker.” “That was educational!” “Still banned from Story Time at the Westside Library.” A squealing child ran past, chasing a deflating balloon shaped like a llama. Bill sighed. “I hate hurting kids.” Molly snapped on rubber gloves. “We’re not hurting them. We’re teaching them about capitalism.” “Still feels a little evil.” “It’s *efficient*. Now move. If I don’t get that remote-control taco, I’m going full Grinch.”
created_at
"2026-03-25T10:18:23.272Z"
3/25/2026, 10:18:23 AM
model
"qwen/qwen3-235b-a22b-07-25"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
68
moderation_latency
(null)
generation_time
3638
tokens_prompt
77
tokens_completion
386
native_tokens_prompt
72
native_tokens_completion
304
native_tokens_completion_images
(null)
native_tokens_reasoning
0
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
"stop"
usage
0.0000376
router
(null)
provider_responses
0
endpoint_id
"c9f927db-d559-49eb-9846-80862c5e3b4e"
id
"chatcmpl-bac154a5b6b010c3"
is_byok
false
latency
31
model_permaslug
"qwen/qwen3-235b-a22b-07-25"
provider_name
"WandB"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1774433903-lLdFGodsdNMt7XxBsMIw"
upstream_id
"chatcmpl-bac154a5b6b010c3"
total_cost
0.0000376
cache_discount
(null)
upstream_inference_cost
0
provider_name
"WandB"
Evaluation details
Result
Evaluator
Details
Meta Data
75.16%
Matches word count
n/a
neededClean
false
words
213
0.00%
Dialogue to Total Word Ratio
Ratio: 80.37%, Deviation: 30.37%
neededClean
false
wordsTotal
214
wordsDialogue
172
37.5778%