Simon Willison·· 9 d ago
Claude Haiku 5.5
Claude Haiku 5.5
AI summary
As previously promised , here's Anthropic's new fast, low cost model: Introducing Claude Haiku 5.5 . The previous Haiku, 4.5, was very much showing its age.
Selection record
Threshold 76Media and individualsFirst 78Second 78
AdmittedSum of both 156 ≥ twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed：介绍Claude新模型及API定价评测
- Why it was chosen
- Benchmarks Haiku 5.5's pricing, its new tokenizer and the cost and latency of each reasoning effort level, so you can judge the trade-offs against GPT-6 Luna for long-context work.
- Same event
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: Simon Willison · simonwillison.net