NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-04-25-01KQ2JCVTRSAMHKJAXJEXCQE9G
Run Details
Model:
DeepSeek V4 Pro
Test:
Dialogue tags
Scenario:
Write 200 words with 10% dialogue
Duration:
11.8s
Cost:
$0.000296
Provider:
DeepSeek
Evaluation summary
Score
Evaluator
Details Link
6.09%
Matches word count
(details)
0.00%
Dialogue to Total Word Ratio
(details)
3.0453%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 200 word scene that has 10% 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 community centre’s cinderblock walls practically sweated with the muffled thump of a pop song about a baby shark. Molly unfolded a crudely drawn blueprint of the party room next to a half-eaten tray of supermarket sushi they’d used as a stakeout cover. Bill, wearing a borrowed blazer with sleeves an inch too short, sucked soy sauce off his thumb and squinted. "Security is an octogenarian clown named Sparkles and a bouncy castle," he murmured. "We’ll need a distraction." Molly tapped a finger on the gift table, a glittering mountain of shrink-wrapped plastic and primary-coloured ribbons. "I’ll engage Sparkles in a deeply philosophical conversation about balloon animals," she whispered back. "You breach the present perimeter." Bill shook his head, a grim expression on his face. "Negative. The clown posed an unexpected psychological threat at the door. He offered to make me a poodle and I experienced a primal fear." Molly sighed, rolling up the blueprint. "Fine. We use the ultimate diversion." She produced a single, industrial-grade party popper from her jacket, its promise of chaos glinting in the fluorescent light. "You pop this near the sugar cookie table," she explained, "and in the ensuing toddler riot, I’ll wade in, a ghost through the anarchy, and liberate the gifts." Bill looked at the humble explosive with newfound respect. "How do we exfil?" he asked.
created_at
"2026-04-25T14:58:45.367Z"
4/25/2026, 2:58:45 PM
model
"deepseek/deepseek-v4-pro-20260423"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1034
moderation_latency
(null)
generation_time
10077
tokens_prompt
77
tokens_completion
350
native_tokens_prompt
67
native_tokens_completion
307
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)
num_fetches
(null)
web_search_engine
(null)
origin
"https://ncbench.com/"
is_byok
false
finish_reason
"stop"
native_finish_reason
"stop"
usage
0.000296235
router
(null)
provider_responses
0
endpoint_id
"d084170b-dd99-4e53-9ae1-b63881f3cfbf"
id
"63184a8b-991a-40a0-a02c-4cf835e4a732"
is_byok
false
latency
283
model_permaslug
"deepseek/deepseek-v4-pro-20260423"
provider_name
"DeepSeek"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.14.1; linux; x64))"
http_referer
(null)
request_id
"req-1777129125-K3ljVYYZnWm7AWyd08at"
session_id
(null)
api_type
"completions"
id
"gen-1777129125-PixAGW9jkcjhYrFfXJHI"
upstream_id
"63184a8b-991a-40a0-a02c-4cf835e4a732"
total_cost
0.000296235
cache_discount
(null)
upstream_inference_cost
0
provider_name
"DeepSeek"
response_cache_source_id
(null)
Evaluation details
Result
Evaluator
Details
Meta Data
6.09%
Matches word count
n/a
neededClean
false
words
223
0.00%
Dialogue to Total Word Ratio
Ratio: 40.09%, Deviation: 30.09%
neededClean
false
wordsTotal
227
wordsDialogue
91
3.0453%