EfficientAnthropic

Claude Haiku 3.5

Context

200K tokens

Pricing

$0.80/M input, $4/M output

Modalities

text, image, code

Released

Oct 2024

Overview
Anthropic's fastest and most cost-effective model, optimized for near-instant responses at high throughput. Claude Haiku 3.5 is designed for latency-sensitive workloads like classification, extraction, and real-time chat.
Why it matters
When you need Claude-level safety and instruction-following at commodity pricing, Haiku 3.5 is the answer. It unlocks use cases where per-token cost was previously prohibitive — think high-volume customer support, document triage, or always-on copilot features embedded in a product. For CTOs, it means you can deploy Claude quality into every tier of your stack without blowing your inference budget. Investors should note that Anthropic's ability to compress capability into cheaper tiers pressures the entire market on price-performance.

Key strengths

  • Sub-second latency for most queries
  • Strong instruction-following at low cost
  • Vision capability included
  • Best cost-per-quality ratio in the Claude family

We cover ai models every week.

Get the 5 AI stories that matter — free, every Friday.

Know the terms. Know the moves.

Get the 5 AI stories that matter every Friday — free.

Free forever. No spam.