BeClaude
Claude 3Released 2024-03-13

Claude 3 Haiku

Most affordable Claude model

Family
Claude 3
Release Date
2024-03-13
Best For
Simple tasks, ultra-high-volume processing, cost-sensitive applications

Overview

Claude 3 Haiku is the most affordable Claude model at just $0.25/M input and $1.25/M output tokens. Designed for high-volume, low-latency applications where cost efficiency is paramount. Maintains the full 200K context window and delivers reliable performance on straightforward tasks.

Pricing

Input / 1M tokens
$0.25
Output / 1M tokens
$1.25
Context Window
200K
Max Output
4K
API model ID: claude-3-haiku-20240307

Key Features

Lowest cost: $0.25/M input, $1.25/M output
Fastest response times in the Claude lineup
Full 200K token context window
Reliable for straightforward tasks
Ideal for cost-sensitive high-volume deployments

Use Cases

Ultra-high-volume text processing
Simple classification and routing
Keyword extraction and basic NLP
Customer support triage
Moderation of user-generated content at scale

Benchmarks

BenchmarkScore
MMLU (5-shot)68.5%
Coding (HumanEval)75.9%