Back to models
Anthropic logo

Anthropic

Claude 3.5 Haiku

Fast and lightweight model ideal for summarization, routing, support tasks, and budget-sensitive assistants.

Overall score
84
fastbudgetsummaries
Context window
200K
Speed
Fast
Input pricing
$0.8 / 1M tok
Output pricing
$4 / 1M tok

Score breakdown

Capability78
Use-case fit85
Cost efficiency92
Speed96
Reliability88
Agent readiness80
Ecosystem85

Scores combine benchmark signals, product experience, and editorial weighting. Use them as a practical guide, not an absolute truth claim.

Best for

researchagent-automation

Works with

LangGraphSerpApiSlack assistants

Modalities

textimage

Strengths

  • Responsive and economical for high-volume operations
  • Strong default tone quality for concise outputs

Things to watch

  • Limited ceiling for complex multi-step planning
  • Less appealing for agent workflows that require richer tool state

Best for