←All models
Directory · Models · Anthropic
Claude 4.xClaude Haiku 4.5
Fast, cheap, and smart enough for most routing and extraction tasks. Great base model for subagents.
Input / 1M
$0.80
Output / 1M
$4.00
Context
200K
Released
Oct 1, 2025
Benchmark profile
Profile
MMLU-Pro81.4%
HumanEval87.0%
SWE-Bench49.2%
MATH76.8%
IFEval87.3%
ARC-AGI76.8%
Scores
Editorial
Strengths
- +Sub-second latency
- +10x cheaper than Opus
- +Strong structured output
Trade-offs
- −Not meant for complex multi-step reasoning
Price history (last 90 days)
Input / 1M
Output / 1M
What this costs to use
All task cards →Recommended alternatives
Where to buy
All providers →Input pricing · $/1M tokens · sorted cheapest first