Claude Haiku 4.5
Mid-TierMultimodalAPI Available
Benchmark Scores
83.4/100MMLU85
HumanEval88.5
Math74.2
Reasoning82.5
Coding87
About Claude Haiku 4.5
Claude Haiku 4.5 is Anthropic's fastest and most affordable model, optimized for high-throughput applications. It delivers impressive performance for its cost tier, making it ideal for production workloads that need speed and efficiency.
Strengths
- +Very fast inference
- +Low cost
- +Strong coding for its tier
- +Good for high-volume tasks
Weaknesses
- -Weaker reasoning than Sonnet
- -Less nuanced creative writing
- -Smaller knowledge base
Pricing
Per 1M tokens
Input$0.80
Output$4.00
Quick Facts
- Context Window
- 200K
- Parameters
- Undisclosed
- Release Date
- 2025-11-15
- Category
- Mid-Tier
Compare Models
See how Claude Haiku 4.5 stacks up against other models