NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-08-01M4DAGT5J0NTFERJXY1NPF4MM
Run Details
Model:
Claude Haiku 5.5 (Adaptive)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
9.0s
Cost:
$0.000705
Provider:
Claude Platform on AWS
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
99.62%
Dialogue to Total Word Ratio
(details)
49.8084%
Overall
Model input
User/Human
You are a author of comedic heist books. Please write a 500 word scene that has 70% 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 van sat in the shadow of the church hall car park, engine off, headlights off, and Bill was wearing a fake ginger moustache that kept sliding toward his chin. "Stop fidgeting," Molly hissed. "You look like a walrus that's been shoplifting." "It's itchy. What's it made of, yak?" "It's a disguise, Bill. Nobody looks twice at a man with a moustache at a children's party." "Nobody looks twice at a man in a paper crown, either, apparently." He tugged at the gold cardboard on his head. "Why am I the king?" "You're the magician's assistant. Magicians need assistants. Assistants carry the bags." Bill peered through the windscreen at the glowing hall. Pink balloons bobbed against the glass. Somewhere inside, a man in a rainbow wig was shrieking the words to a song about a tortoise. "Right. Run through it again," Bill said. "Twenty-three presents, all in the side room, all wrapped in that shiny paper." "Twenty-three presents, one big pile, and the parents are in the kitchen drinking prosecco and pretending they're not listening to the clown's divorce story." "Why is the clown telling them about his divorce?" "Because the clown is their cousin, and he's been drinking the prosecco too." "Fine. And the dog?" "What dog?" "The dog in your notes, Molly. The one with the little jumper. You wrote 'beware the dog' and underlined it three times." "Oh. That's Biscuit. He's harmless. Mostly." "Mostly?" "He only bites people who mention the word 'vet.'" Bill slumped. "This was supposed to be a glamorous job." "Stealing from seven-year-olds is glamorous, Bill. Think of the look on their faces." "That's exactly what worries me. Last time, a little girl chased us down the road with a plastic sword." "She had good form. Now, you go in first, carry the big cardboard box marked 'magic props,' and head straight for the side room. I'll follow with the bag. If anyone asks, you're fixing the disco lights." "And if anyone asks what's in the box?" "Tell them it's the rabbit." "There's no rabbit." "Exactly. Nobody wants to see a rabbit that isn't there. Let's go." Bill pulled open the door, took a deep breath, and stepped out into the cold, a paper crown glinting on his head like the world's least convincing king. Behind him, Molly grinned and slung the empty sack over her shoulder.
created_at
"2026-10-08T08:36:46.906Z"
10/8/2026, 8:36:46 AM
model
"anthropic/claude-haiku-5.5-20261007"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
8965
moderation_latency
(null)
generation_time
8965
tokens_prompt
75
tokens_completion
592
native_tokens_prompt
95
native_tokens_completion
1390
native_tokens_completion_images
(null)
native_tokens_reasoning
529
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
"end_turn"
service_tier
"default"
usage
0.0007045
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791448606-yNQv4S3fLFdjP0MLV5IW"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791448606-YbdJXaT64wVP6d1xy0zg"
upstream_id
"msg_011CfpTxCknywgkbJSa5Be1R"
provider_responses
0
endpoint_id
"b6cdf493-3a61-441f-a134-e7944fddb980"
id
"msg_011CfpTxCknywgkbJSa5Be1R"
is_byok
false
latency
592
model_permaslug
"anthropic/claude-haiku-5.5-20261007"
provider_name
"Claude Platform on AWS"
status
200
total_cost
0.0007045
cache_discount
(null)
upstream_inference_cost
0
provider_name
"Claude Platform on AWS"
response_cache_source_id
(null)
data_region
"global"
workspace_id
"97e315e5-d303-487d-83c1-83180e8a13d4"
Evaluation details
Result
Evaluator
Details
Meta Data
0.00%
Matches word count
n/a
neededClean
false
words
393
99.62%
Dialogue to Total Word Ratio
Ratio: 70.28%, Deviation: 0.28%
neededClean
false
wordsTotal
397
wordsDialogue
279
49.8084%