NC Bench
Overview
About
Tests
Models
Vendors
Run: 2026-02-16-01KHM5FSBF5CWXGRM4Q8C2ZPDW
Run Details
Model:
DeepSeek V3 (2024-12-26)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
13.2s
Cost:
$0.000377
Provider:
DeepInfra
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 90% 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
“Bill, listen, this is foolproof,” Molly whispered, crouched behind a hedge outside the brightly lit backyard. “Foolproof? It’s a kid’s birthday party, Molly. There’s clown balloons and a bounce house!” “Exactly! Distraction city. No one’s gonna notice two ‘party planners’ slipping out with a few gifts.” “A few gifts? You said we were taking *all* the presents.” “Details, Bill. Focus. You’re dressed as the piñata guy. I’m the caterer. We blend in, we grab, we go.” “Why am I the piñata guy? That’s humiliating.” “Because you’re the only one who fits in the costume. Now, stick to the plan. You distract the kids by letting them hit you with sticks—” “Wait, what? They’re gonna hit me?” “It’s a piñata, Bill. It’s part of the job.” “You didn’t say anything about getting hit!” “Relax! It’s padded. Meanwhile, I’ll sneak into the gift room, load ’em up, and meet you by the trampoline.” “What if someone catches us?” “Who’s gonna catch us? A six-year-old with a juice box? C’mon, Bill, show some guts.” “Fine. But if I get frosting in my hair, you owe me dessert.” “Deal. Now go! The clown’s juggling flamingos—it’s our moment.” Bill sighed, pulling the piñata costume over his head. “This better be worth it.” Molly grinned. “Trust me. Worst case, we leave with cake.” “Cake isn’t loot, Molly.” “It is if you’re desperate.”
created_at
"2026-02-16T21:25:31.154238+00:00"
2/16/2026, 9:25:31 PM
model
"deepseek/deepseek-chat-v3"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
302
moderation_latency
(null)
generation_time
13202
tokens_prompt
77
tokens_completion
423
native_tokens_prompt
66
native_tokens_completion
400
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
0.00037712
router
(null)
provider_responses
0
id
"chatcmpl-RX9KfPFc5U5eYJ1jDuABydfF"
status
200
is_byok
false
latency
39
endpoint_id
"5294d55f-9012-496b-8f22-8cc919432dcd"
provider_name
"DeepInfra"
model_permaslug
"deepseek/deepseek-chat-v3"
api_type
"completions"
id
"gen-1771277116-kiUUwa3iGRQE3BdlOUzk"
upstream_id
"chatcmpl-RX9KfPFc5U5eYJ1jDuABydfF"
total_cost
0.00037712
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepInfra"
Evaluation details
Result
Evaluator
Details
Meta Data
0.0848%
Matches word count
n/a
neededClean
false
words
229
99.9998%
Dialogue to Total Word Ratio
Ratio: 90.39%, Deviation: 0.39%
neededClean
false
wordsTotal
229
wordsDialogue
207
50.0423%