NC Bench
Overview
About
Tests
Categories
Models
Vendors
Compare
Run: 2026-10-08-01M4DA22WY406R13J5QQZSHNZK
Run Details
Model:
Claude Haiku 5.5 (Reasoning, Medium)
Test:
Dialogue tags
Scenario:
Write 500 words with 70% dialogue
Duration:
6.1s
Cost:
$0.000427
Provider:
Claude Platform on AWS
Evaluation summary
Score
Evaluator
Details Link
0.00%
Matches word count
(details)
25.62%
Dialogue to Total Word Ratio
(details)
12.8120%
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
Molly crouched behind the hedge, squinting at the bright balloons bobbing above the back garden. "Right, Bill. Talk me through the plan again." Bill adjusted his fake moustache, which kept drooping at one end. "Simple. We wait for the clown to finish the balloon animals. Then we slip through the side gate, grab every present with a bow on it, and vanish." "Every present?" "Every single one. Those kids won't even notice until they've had their cake." Molly peered through the gap in the fence. A small boy in a paper crown was jumping on a bouncy castle. "That one's going to notice. He's got his eye on the pile by the patio." "Nobody notices a bouncy castle, Molly. That's the whole point of a bouncy castle." "Fair point." She tugged at the sleeve of her black jumper. "Where did you get this bag, anyway? It's got teddy bears printed on it." "It was the only one left at the charity shop. Look, it's practically invisible." "It is literally a bag covered in teddy bears, Bill." "Okay, it's not invisible. But it's cheap, and we're not exactly cutting costs on this job." Molly sighed. "How many presents do we need?" "Thirty-seven. I counted them through the window this morning." "Thirty-seven? Bill, that's a lot of paper and tape. What if they're all just socks?" "Then we'll have very warm feet and a grudge against the birthday girl." A whistle sounded from inside the house. Molly grabbed his arm. "That's the signal. The clown's doing the sausage dog balloon. Go, go, go!" They scrambled over the low wall and sprinted across the lawn, Bill's moustache finally giving up and flapping onto his cheek. Molly seized the first gift, a box wrapped in glittery pink paper. "Oh, this one's heavy," she whispered. "Probably a puppy." "Bill, don't say that." "Or a very large book." They reached the gate just as a tiny voice shrieked behind them, "Mummy, the presents are walking!" "Run!" Molly hissed, stuffing a teddy-bear bag under her arm. "Keep calm," Bill panted. "We've got this. Probably." "You said probably last time, and we ended up in a ball pit." "That was a different party." Molly pulled the gate shut behind them and grinned. "Right. Where to next?" "Birthday party number two. Same clown, different cake."
created_at
"2026-10-08T08:28:44.326Z"
10/8/2026, 8:28:44 AM
model
"anthropic/claude-haiku-5.5-20261007"
app_id
182717
external_user
(null)
streamed
true
cancelled
false
latency
755
moderation_latency
(null)
generation_time
6114
tokens_prompt
75
tokens_completion
577
native_tokens_prompt
95
native_tokens_completion
835
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
"end_turn"
service_tier
"default"
usage
0.000427
router
(null)
user_agent
"langchainjs-openai/1.0.0 ((node/v24.21.0; linux; x64))"
http_referer
(null)
request_id
"req-1791448124-I5BFvjS1NLAH7QFBJ1Fh"
session_id
(null)
preset_id
(null)
api_type
"completions"
id
"gen-1791448124-6eOIyMK3l2JcDiNELri8"
upstream_id
"msg_011CfpTLdregATo1cM6Eh6Gb"
provider_responses
0
endpoint_id
"b6cdf493-3a61-441f-a134-e7944fddb980"
id
"msg_011CfpTLdregATo1cM6Eh6Gb"
is_byok
false
latency
755
model_permaslug
"anthropic/claude-haiku-5.5-20261007"
provider_name
"Claude Platform on AWS"
status
200
total_cost
0.000427
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
386
25.62%
Dialogue to Total Word Ratio
Ratio: 64.78%, Deviation: 5.22%
neededClean
false
wordsTotal
389
wordsDialogue
252
12.8120%