Loading page
LightweightAnthropic
Claude 3 Haiku
Claude 3 Haiku is Anthropic's lightweight model designed for fast, cost-effective tasks with text and image support and a 200K token context window.
Context 200K
Tier Lightweight
Modalities text, image
Input from
$0.125 / 1M tokens
across 2 providers
API Pricing
Cheapest on Amazon AWS — 40% below avg| Provider | Input / 1M | Output / 1M | Cached / 1M |
|---|---|---|---|
| $0.125 | $0.625 | - | |
| $0.250 | $1.25 | - | |
| $0.250 | $1.25 | $0.030 |
Prices updated daily. Last check: Sep 6, 2026
Claude 3 Haiku pricing by provider
Input, output, and batch rates, plus alternatives, for one provider at a time.
Performance & Benchmarks
Source: Artificial Analysis →Intelligence
1.0 / 100
Reasoning & Knowledge
- GPQA Diamond37.4%
- Humanity's Last Exam4.1%
Coding
- LiveCodeBench15.4%
Math
- AIME1.0%
- MATH-50039.4%
Agentic & Tool Use
- Terminal-Bench Hard0.8%
- τ²-bench21.1%
Instruction & Long Context
- IFBench36.1%
- Long-Context Reasoning27.7%
Benchmarks measured Sep 2026. Scores are independent evaluations, not vendor-reported.
Model Details
General
- Creator
- Anthropic
- Family
- Haiku
- Tier
- Lightweight
- Context Window
- 200K
- Modalities
- Text, Image
Capabilities
- Tool Calling
- No
- Open Source
- No
- Aliases
- Haiku 3
Strengths & Limitations
Strengths
- Fast response times with 130.52 output tokens per second
- Low latency with 530ms time to first token
- Supports both text and image inputs for multimodal tasks
- Large 200K token context window for processing lengthy documents
- Optimized for cost-effective high-volume applications
- Maintains Anthropic's constitutional AI safety approach
- Good performance on basic reasoning and writing tasks
Limitations
- No function calling or tool use capabilities
- Proprietary model with no open-source weights available
- Less capable than newer Claude 3.5 Sonnet and Claude 3 Opus models
- Limited advanced reasoning compared to flagship-tier models
- No structured output modes like JSON formatting
Key Features
•200K token context window
•Text and image input support
•High-speed inference (130.52 tokens/second)
•Low latency responses (530ms TTFT)
•Streaming response capability
•Constitutional AI safety training
•Batch processing support
•API integration with major cloud providers
About Claude 3 Haiku
Claude 3 Haiku is Anthropic's lightweight model in the Claude 3 family, positioned as the fastest and most economical option for high-volume applications. While Anthropic has released newer flagship models like Claude 3.5 Sonnet and Claude 3 Opus, Haiku remains the go-to choice for tasks requiring speed and efficiency over maximum capability.
The model supports both text and image inputs with a 200K token context window, making it suitable for processing lengthy documents and visual content. Haiku delivers 130.52 output tokens per second with a 530ms time to first token, emphasizing rapid response times. The model handles basic reasoning, writing, and analysis tasks while maintaining Anthropic's safety standards.
Claude 3 Haiku is commonly used for content moderation, customer support automation, data extraction from documents, and other high-throughput applications where speed and cost efficiency matter more than advanced reasoning capabilities. It competes with other lightweight models like GPT-3.5 Turbo in scenarios requiring quick turnaround times for simpler tasks.
Common Use Cases
Claude 3 Haiku excels at high-volume, cost-sensitive applications where speed matters more than advanced reasoning. Common use cases include content moderation at scale, customer support chatbots, document summarization and data extraction, basic writing assistance, and simple classification tasks. Its multimodal capabilities make it suitable for processing mixed text and image content in workflows like document analysis, basic visual question answering, and image captioning. The large context window allows it to handle lengthy documents, transcripts, and conversations while maintaining fast response times for real-time applications.
Frequently Asked Questions
How much does Claude 3 Haiku cost per million tokens?
Claude 3 Haiku pricing varies by provider and pricing type (standard vs batch processing). Check the pricing table above for current rates across all providers offering this model.
What is Claude 3 Haiku best used for?
Claude 3 Haiku is designed for high-volume, cost-sensitive applications requiring fast responses. It works well for content moderation, customer support automation, document summarization, basic writing tasks, and simple classification. Its speed and multimodal capabilities make it ideal when you need quick processing of text and images without requiring advanced reasoning.
Does Claude 3 Haiku support function calling and tool use?
No, Claude 3 Haiku does not support function calling or tool use capabilities. For applications requiring tool integration, you would need to use Claude 3.5 Sonnet or other models in Anthropic's lineup that include these features.