Anthropic Unveils Claude Haiku 4.5, a Faster and Smarter AI Model for Develop
Anthropic has introduced Claude Haiku 4.5, the latest version of its compact AI model. It offers comparable performance to Sonnet 4 but runs at one-third the cost and more than twice the speed. The model, launched on Wednesday, demonstrates how smaller systems can deliver impressive capabilities without the heavy infrastructure demands of larger models.
According to internal benchmarks, Haiku scored 73% on SWE-Bench verified and 41% on Terminal-Bench. While these results fall slightly below Sonnet 4.5, they remain competitive with Sonnet 4, GPT-5, and Gemini 2.5. Similar scores appear across tool use, computer use, and visual reasoning benchmarks. These outcomes confirm that Haiku balances efficiency and power effectively.
The model is now available under all free plans and is designed for wide accessibility. Its lightweight design allows multiple Haiku agents to run simultaneously or alongside more advanced systems. As a result, it can significantly reduce server load while maintaining strong performance.
Expanding AI Deployment Possibilities
Mike Krieger, Chief Product Officer, emphasized that Haiku 4.5 makes new styles of AI deployment practical in production environments for the first time. “It’s opening up entirely new categories of what’s possible with AI in production environments – with Sonnet handling complex planning while Haiku-powered sub-agents execute at speed,” Krieger said. “We’re giving people a complete agent toolbox where each model has the right combination of intelligence, speed, and cost for different parts of the job.”
This development demonstrates how combining models of varying scales can improve both flexibility and output. Furthermore, it paves the way for AI systems that adapt dynamically to specific workloads. Developers can integrate Haiku 4.5 for quick, task-oriented execution while reserving larger models for strategic or analytical work. Consequently, teams gain more control over performance, latency, and overall costs.
Ideal for Developers and Fast-Response Tools
Haiku 4.5 is expected to find its strongest use in software development tools, where speed and responsiveness are essential. With Claude Code already widely adopted, the new model can further enhance productivity by minimizing latency. Andrew Filev, CEO of Zencoder, noted that Haiku 4.5 is “unlocking an entirely new set of use cases.”
The launch follows two recent releases: Sonnet 4.5, introduced two weeks earlier, and Opus 4.1, released two months ago. Each model has advanced the company’s broader ecosystem, offering scalable solutions that cater to different user needs. The previous Haiku version debuted in October 2024, marking nearly a year of rapid innovation.
Haiku 4.5 strengthens the company’s lineup, ensuring that users gain access to faster, cost-effective AI capabilities suited for both experimentation and large-scale deployment.
