NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-03-25-01KMJ7Z9J68EJB42X4D130P99S
Run Details
Model:
Qwen3 235B A22B Instruct 2507
Test:
Dialogue tags
Scenario:
Write 200 words with 50% dialogue
Duration:
20.6s
Cost:
$0.000201
Provider:
SiliconFlow
Evaluation summary
Score
Evaluator
Details Link
99.38%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
49.6885%
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 crouched behind a shrub, binoculars glued to her eyes. “It’s worse than I thought. Three piñatas, two bounce houses, and—oh no—a clown. A *real* clown.” Bill adjusted his backpack full of duffel bags. “Clown’s not even armed. What’s the worst he can do? Honk me to death?” “He’s holding a chainsaw-shaped lollipop, Bill. That’s not a snack. That’s a psychological weapon.” “So we time it between songs. When they’re all screaming 'Baby Shark,' we’ll move in. Distraction level: maximum.” “Wait—there’s a *parrot* on the clown’s shoulder. Birds ruin everything. Remember Miami? One squawk and the whole operation went sideways.” “Parrots don’t have thumbs, Molly. They can’t dial 911.” “They can peck. And make noise. And possibly remember my face.” “Fine. We use Operation Squirrel. You cause a distraction with the glitter bomb. I grab the loot.” “That glitter bomb blinded us for three days last time.” “It added *pizzazz*.” “This is a children’s party, not a stage show.” “We’re professionals. We don’t care if they’re six or sixty. If it’s wrapped, it’s ours.” “What about the cake?” “Only as a last resort. Frosting slows us down.” “Bill, are we evil?” “Only between 7 PM and 7:05. After that, we’re just misunderstood entrepreneurs.”
created_at
"2026-03-25T10:17:18.665Z"
3/25/2026, 10:17:18 AM
model
"qwen/qwen3-235b-a22b-07-25"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1506
moderation_latency
(null)
generation_time
20561
tokens_prompt
77
tokens_completion
379
native_tokens_prompt
72
native_tokens_completion
324
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
0.00020088
router
(null)
provider_responses
0
endpoint_id
"ce835ea8-00c7-4686-831d-7733dac5e2d9"
id
"019d247fa945be4ca4fc7ba4ef02b857"
is_byok
false
latency
1506
model_permaslug
"qwen/qwen3-235b-a22b-07-25"
provider_name
"SiliconFlow"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.0; linux; x64))"
http_referer
(null)
api_type
"completions"
id
"gen-1774433838-5CJugbXqDtE9sefMevN1"
upstream_id
"019d247fa945be4ca4fc7ba4ef02b857"
total_cost
0.00020088
cache_discount
(null)
upstream_inference_cost
0
provider_name
"SiliconFlow"
Evaluation details
Result
Evaluator
Details
Meta Data
99.38%
Matches word count
n/a
neededClean
false
words
205
0.00%
Dialogue to Total Word Ratio
Ratio: 91.26%, Deviation: 41.26%
neededClean
false
wordsTotal
206
wordsDialogue
188
49.6885%