Post

Anthropic just released Haiku 5.5, and it's quite cheap, matching GPT-6 Luna pricing for short context. Long context is 5x the cost though (2x for Luna). Previous reporting by Epoch has suggested that larger Claudes may use linear/sparse attention, but this pricing structure indicates Haiku may not.

Here's the previous reporting by Epoch:

epoch.aiLatency scaling differences for GPT and Claude modelsEpoch AI measures long-context latency scaling across four frontier models. GPT-5.6 Terra and Sol show quadratic time-to-first-token growth with context length, while Claude Sonnet 5 and Opus 5 stay n...

7/10/26, 6:50 pm0 reposts5 likes

Here's the Haiku 5.5 release page:

anthropic.comIntroducing Claude Haiku 5.5Claude Haiku 5.5 is our fastest, most capable small model. Built for high-volume work like summarization, subagents, and browser use.