/ Models Index / Claude Haiku 4.5

Claude Haiku 4.5

Frontier · Anthropic · Anthropic · Released Q1 2026
A-

Excellent (A-) — Capability Grade

Claude Haiku 4.5 is Anthropic's small-model frontier offering. Delivers Sonnet-class reasoning on most non-agentic benchmarks at a fraction of the cost. Strong cost-efficient tier for high-throughput production workloads. The fastest Claude model in latency-sensitive applications.

89
Composite / 100
/ Subscore Breakdown · 6 Capability Dimensions

Where this grade comes from.

General Reasoning
B+
Code Generation
B+
Math & STEM
B+
Tool Use & Agency
B+
Multimodal
B+
Safety & Alignment
A-
/ Key Events & Disclosures

Release timeline & positioning.

  • Released Q1 2026
  • Cost-efficient tier
  • Strong production latency
  • Anthropic small-model frontier

/ Best for

High-throughput production workloads requiring Claude-quality reasoning at the lowest Claude-family price point.

/ Watch out for

Agentic and tool-use depth lags Sonnet 4.6 and Opus 4.6. Reserve Haiku for batch reasoning vs. complex multi-step workflows.