Claude Haiku 4.5

Provider: Anthropic US Official API price

Input price
$1.00
/ 1M tokens
Output price
$5.00
/ 1M tokens
Cache price
$0.10
/ 1M tokens (read)
Batch API
$0.50 / $2.50
input / output per 1M tokens
Context
200K tokens
Max output 64K tokens

Overview

Anthropic's Claude Haiku 4.5 is a cost-optimized model designed for efficient performance across a range of applications. It belongs to the Haiku tier, known for its balance of capability and affordability while offering advanced reasoning. This model is suitable for general-purpose workloads requiring quick responses and solid reasoning capabilities.

Strengths

  • Provides an extremely large context window of 200,000 tokens for extensive data processing.
  • Supports vision capabilities, enabling the analysis and understanding of image-based content.
  • Features function-calling and JSON mode for structured outputs and robust tool integration.
  • Offers strong reasoning capabilities, making it suitable for complex analytical tasks.
  • Generous maximum output length of 64,000 tokens allows for detailed and comprehensive responses.

Best use cases

  • Summarizing or analyzing very long documents, codebases, or extensive chat logs.
  • Processing and extracting information from various image formats and visual data.
  • Building intelligent agents or tools that require structured outputs and external API interaction.
  • Developing applications needing reliable reasoning for complex problem-solving and analysis.
  • Generating extensive reports, detailed creative content, or long-form documentation efficiently.

Pricing position

Claude Haiku 4.5 is priced at $1/1M input tokens and $5/1M output tokens, positioning it above earlier Haiku models like Claude 3 Haiku ($0.25/$1.25 per M tokens) and Claude 3.5 Haiku ($0.80/$4 per M tokens). While its input price aligns with peers such as Perplexity's Sonar and xAI's Grok Build, its output price of $5/1M tokens is higher than many, though less than Qwen3.6 Max Preview's $6.24/1M output tokens. It remains significantly more affordable than the Anthropic Sonnet series, which starts at $3/1M input tokens and $15/1M output tokens.

Caveats

Claude Haiku 4.5 does not offer a free tier for developers to experiment with without cost. For applications that do not require its extensive 200,000-token context window or vision capabilities, previous Haiku versions may offer a more cost-effective solution. Its higher output token pricing compared to some peers could be a consideration for output-heavy applications.

Notes drafted by an LLM from the figures on this page (2026-05-21) and rewritten when the price changes.

Key facts

Price basis
Official API price
Max output
64K tokens
Listed on OpenRouter
2025-10-15
Tracked since
2026-05-07
Last updated
2026-11-10

Features

Image inputFunction callingJSON outputReasoning

Monthly cost estimate

Scale Input (M) Output (M) Monthly
0.1M (hobby) 0.1 0.05 $0.35
1M (small app) 1 0.5 $3.50
10M (mid scale) 10 5 $35.00
100M (large) 100 50 $350.00

※ Cache and batch discounts not factored in. Detailed calculator

Official resources

Other models from Anthropic

Similar-priced models

Source: OpenRouter API (anthropic/claude-haiku-4.5)