Anthropic has launched Claude Haiku 4.5, its latest small and efficient AI model. The company states that the new model delivers performance comparable to its Sonnet 4 model, which was considered state-of-the-art just five months prior. According to Anthropic, Haiku 4.5 achieves this at one-third the cost and more than twice the speed of Sonnet 4. The company supports these claims with benchmarks showing it matches or exceeds Sonnet 4 in areas like coding and computer use tasks.
In a significant move, Anthropic is making Haiku 4.5 available to all free users of its Claude.ai platform, democratizing access to what it calls “near-frontier” AI. The model’s speed makes it ideal for real-time applications like customer service chats and responsive coding assistants.
A new multi-agent approach
Anthropic also introduced a new method for tackling complex problems. A more powerful model, like Sonnet 4.5, can plan and break down a large project into smaller steps. It then orchestrates a team of multiple Haiku 4.5 models to execute these sub-tasks in parallel. This approach, inspired by how human teams work, can accelerate large software projects or research analysis.
Anthropic classified the model under its AI Safety Level 2 (ASL-2) standard, which is less restrictive than for its larger models. The company explained this is based on extensive safety testing. Developers can now use Claude Haiku 4.5 through Anthropic’s API and on cloud platforms from Amazon and Google.
Additional sources: VentureBeat, TechCrunch
