Anthropic · Claude 4

Claude Haiku 4.5

Fast, cheap, surprisingly capable for high-volume jobs.

Claude Haiku 4.5, developed by Anthropic and released in 2023, is a flagship multimodal AI model within the Claude 4 family. It is designed for professional use cases, combining robust natural language processing with the ability to handle both text and image inputs effectively. The model is particularly suited for developers and product teams, providing capabilities for complex workflows and seamless multimodal interaction.

Claude Haiku 4.5 is technically distinguished by its large 200,000-token context window, enabling it to maintain coherence over lengthy inputs or conversations. Its training emphasizes safety, alignment, and interpretability, leveraging reinforcement learning from human feedback (RLHF) to ensure reduced harmful outputs and enhanced response quality across diverse applications.

Positioned as a flagship model in the Claude 4 lineup, Claude Haiku 4.5 introduces significant improvements in multimodal processing and context window expansion compared to its predecessors, making it ideal for demanding workflows requiring long-term coherence and seamless text-image integration.

Background

Claude is a series of large language models developed by Anthropic and first released in 2023. Since Claude 3, each generation has typically been released in three sizes, from least to most capable: Haiku, Sonnet, and Opus. An additional model named Claude Mythos was released to some companies in 2026 but not to the public.

Wikipedia

Specs

Context window: 200K tokens
Max output: 64K tokens
Input ($/1M tokens): $1.00
Output ($/1M tokens): $5.00
Modalities: Image · Text
Released: Oct 1, 2025
Weights: Closed

Pricing last synced Apr 27, 2026 via OpenRouter. Confirm against official docs before committing.

Capabilities

Tool use
Vision
Extended thinking
Prompt caching
Open weights

What it excels at

Extensive context window
Enables processing up to 200,000 tokens, supporting long documents and detailed interactions.
Multimodal input processing
Handles both text and image data seamlessly for integrated workflows.
Alignment-focused training
Training with RLHF reduces risks of harmful or disallowed content.
High response coherence
Refined training ensures contextually relevant and detailed outputs over extended interactions.

When to use this model

→Codebase analysis and refactoring — Supports large-scale code review and modifications with its extensive context window.
→Content creation across modalities — Seamlessly generates multimodal content integrating text and image.
→Complex chatbot development — Provides reliable handling of long, multi-turn dialogues for advanced user support systems.
→Research synthesis — Processes and summarizes large volumes of research data effectively.

Analysis synthesized from gpt-4o, llama-4-maverick, phi-4, etc.

API model id

claude-haiku-4-5-20251001

Vendor docs: docs.anthropic.com

Compare Claude Haiku 4.5 with