NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-07-25-01KYCYZ6BJE9ZXPF9HXRH9X0DC
Run Details
Model:
Thinking Machines Inkling (Reasoning)
Test:
Dialogue tags
Scenario:
Write 200 words with 90% dialogue
Duration:
1m 7s
Cost:
$0.013535
Provider:
DeepInfra
Evaluation summary
Score
Evaluator
Details Link
51.93%
Matches word count
(details)
50.57%
Dialogue to Total Word Ratio
(details)
51.2489%
Overall
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
"It's perfect," Molly said. "Thirty kids, one bouncy castle, zero adults who understand eye contact." "We'd be monsters." "We're entrepreneurs. There's a thin, glitter-covered difference." "The party's at noon. I have brunch." "Cancel it. Or bring it. I'm flexible." "You're asking me to rob a six-year-old's unicorn summit." "I'm asking you to redistribute glitter capitalism." "That isn't real." "It'll be on my cards. Listen, the presents are stacked like cardboard Jenga in the corner. I scoped it." "You scouted a birthday party?" "I wore a 'World's Okayest Aunt' shirt and brought juice. I'm practically family." "You drank it?" "I'm not a savage, Bill. I swapped it for espresso. The parents are distracted by a magician who knows one trick, and it's questionable." "Which trick?" "Making dignity vanish. Like ours, if we do this right." "We'd steal from children." "We'd steal from parents who bought robotic dinosaurs forgotten by Tuesday." "Tuesday's optimistic." "Exactly. We accelerate disappointment." "And the castle?" "Distraction. You go high, I go low, we leave with backpacks full of regret." "This feels wrong." "So does your hair, but I stay quiet." "That was once." "Noon. Wear black. Or pastels. Blend with the terror." "The terror?" "The children, Bill. Beautiful, sticky, loud terror. Now pack a sack that says 'Happy Birthday' outside and 'Sorry' inside."
created_at
"2026-07-25T15:40:44.282Z"
7/25/2026, 3:40:44 PM
model
"thinkingmachines/inkling-20260715"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
1120
moderation_latency
(null)
generation_time
66283
tokens_prompt
75
tokens_completion
3346
native_tokens_prompt
75
native_tokens_completion
3330
native_tokens_completion_images
(null)
native_tokens_reasoning
2994
native_tokens_cached
32
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
0.01353494
router
(null)
provider_responses
0
endpoint_id
"d6b9d68e-115a-49d7-81cd-1ad1c6942b0e"
id
"chatcmpl-Ru2KHmiHHchW6mPCnyPA6LdX"
is_byok
false
latency
258
model_permaslug
"thinkingmachines/inkling-20260715"
provider_name
"DeepInfra"
status
200
user_agent
"langchainjs-openai/1.0.0 ((node/v24.18.0; linux; x64))"
http_referer
(null)
request_id
"req-1784994044-QGHYCPyvQmBR5rRHhJZI"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1784994044-WcyzRmiMDX159ad6Ky1v"
upstream_id
"chatcmpl-Ru2KHmiHHchW6mPCnyPA6LdX"
total_cost
0.01353494
cache_discount
0.00002656
upstream_inference_cost
0
provider_name
"DeepInfra"
response_cache_source_id
(null)
data_region
"global"
Evaluation details
Result
Evaluator
Details
Meta Data
51.93%
Matches word count
n/a
neededClean
false
words
216
50.57%
Dialogue to Total Word Ratio
Ratio: 99.09%, Deviation: 9.09%
neededClean
false
wordsTotal
219
wordsDialogue
217
51.2489%