NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-09-03-01M1KBW91BS0T711399AFKQM83
Run Details
Model:
DeepSeek V4 Flash 0731 (Reasoning, High)
Test:
Dialogue tags
Scenario:
Write unattributed dialogue
Duration:
15.2s
Cost:
$0.000280
Provider:
Mancer 2
Evaluation summary
Score
Evaluator
Details Link
100.00%
Count dialogue tags
(details)
Model input
User/Human
You are an expert novel writer in the cozy mystery genre. Write a scene of about 200 words, with dialogue between two characters. The dialogue should be unattributed, meaning the speaker is not explicitly mentioned. The scene should be set in a small town bakery. Only return the story, nothing else.
Model response
Messages
Meta
AI
The bell above the bakery door chimed, and the scent of cinnamon and warm sourdough curled through the air. Behind the glass counter, a woman with flour-dusted fingers paused mid-fold of a croissant. “You’re early. The morning batch isn’t even out of the oven yet.” “Early bird catches the, ah, the burnt edge of a scone, I suppose.” He tapped the glass. “That one there. The lopsided one. Looks like it’s been through something.” “They all have stories. That one’s a rum cake. No, wait. That’s a danish. I’m still waking up.” “You’re never still waking up. You’re the one who told me the secret to a good crumb is patience and spite.” “And you’re the one who told me the mayor’s wife didn’t actually slip on the ice last week.” A pause. The woman wiped her hands on her apron, eyes narrowing. “You think she was pushed.” “I think somebody wanted her off the zoning board before the vote on the old mill. And I think you baked her a sympathy pie with a note that said ‘Get well soon’ and your phone number on it.” “I baked her a pie because her husband’s a terrible cook and she’s a good woman.” “You baked her a pie because you wanted to see if she’d mention the shoe print.” The woman smiled slowly. “The shoe print in the snow? The one the sheriff said was a deer?” “The one the sheriff said was a deer,” he echoed. “You’re a terrible liar, Mabel.” “I’m an excellent liar. I just don’t waste it on people who already know the truth.” She slid the lopsided pastry into a bag. “On the house. Consider it a bribe.”
created_at
"2026-09-03T10:08:52.283Z"
9/3/2026, 10:08:52 AM
model
"deepseek/deepseek-v4-flash-20260731"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1323
moderation_latency
(null)
generation_time
13495
tokens_prompt
78
tokens_completion
562
native_tokens_prompt
144
native_tokens_completion
512
native_tokens_completion_images
(null)
native_tokens_reasoning
137
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
0.00028
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.19.0; linux; x64))"
http_referer
(null)
request_id
"req-1788430132-TCJwZjtFNZCd64RZQdKn"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1788430132-D4lWQaYIGXjK9202TTOD"
upstream_id
"01a066be2bcf0091"
provider_responses
0
endpoint_id
"47a93b18-a9c6-485b-9a47-2afb6ca790dc"
is_byok
false
latency
45
model_permaslug
"deepseek/deepseek-v4-flash-20260731"
provider_name
"BaseTen"
status
429
1
endpoint_id
"14c4df90-c7af-42ff-99e7-69d62a89e7d5"
is_byok
false
latency
1188
model_permaslug
"deepseek/deepseek-v4-flash-20260731"
provider_name
"AkashML"
status
429
2
endpoint_id
"fc1cac03-9834-4ed8-8dfe-2dfb8543b690"
is_byok
false
latency
384
model_permaslug
"deepseek/deepseek-v4-flash-20260731"
provider_name
"Morph"
status
429
3
endpoint_id
"9c83b6fb-0614-4d50-b30b-123a09ba354a"
id
"01a066be2bcf0091"
is_byok
false
latency
1236
model_permaslug
"deepseek/deepseek-v4-flash-20260731"
provider_name
"Mancer 2"
status
200
total_cost
0.00028
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Mancer 2"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
100.00%
Count dialogue tags
n/a
neededClean
false
dialogueTags
(empty)