On Wednesday, Anthropic released Claude Haiku 4.5, a small AI language model that reportedly delivers performance similar to what its frontier model Claude Sonnet 4 achieved five months ago but at one-third the cost and more than twice the speed. The new model is available now to all Claude app, web, and API users.
If the benchmarks for Haiku 4.5 reported by Anthropic hold up to independent testing, the fact that the company can match some capabilities of its cutting-edge coding model from only five months ago (and GPT-5 in coding) while providing a dramatic speed increase and cost cut is notable.
As a recap, Anthropic ships the Claude family in three model sizes: Haiku (small), Sonnet (medium), and Opus (large). The larger models are based on larger neural networks and typically include deeper contextual knowledge but are slower and more expensive to run. Due to a technique called distillation, companies like Anthropic have been able to craft smaller-sized AI models that match the capability of larger, older models at functional tasks like coding, although it typically comes at the cost of omitting stored knowledge.
