@codenameposhan: The price is the headline: $0.10 per million input tokens and $0.50 output for prompts up to 100K, a tenth of Haiku 4.5. Anthropic estimates about 75% cheaper on average. The jumps over Haiku 4.5 in Anthropic's evals: OSWorld computer use 15.7% to 72.4%, Terminal-Bench 4.0 0% to 39.2%. And it's the first Haiku with an effort setting. The honest part: Sonnet 5.5 still wins on hard agentic coding (70.6% vs 39.2%), and every score is Anthropic-run. Also today: Sonnet 5.5 cache reads halved, and Max and Team plans get $100 to $500 a month in API credits. I'd use it as the subagent under Sonnet or Opus, not the lead. Follow @codenameposhan for more AI breakdowns. #ClaudeHaiku #Claude #Anthropic #ClaudeCode #AIEngineering #DeveloperTools

Codename Poshan

Open In TikTok:

Region: US

Wednesday 07 October 2026 19:03:01 GMT