Claude Haiku 4.5 Released
Introducing Claude Haiku 4.5
Official announcement (via Hacker News)
Anthropic has released Claude Haiku 4.5 — the most affordable member of the Claude 4.5 family, following the launch of Sonnet 4.5 two weeks ago.
---
Pricing Overview
- Claude Haiku 4.5
- $1 per million input tokens
- $5 per million output tokens
- Compared to previous models:
- Haiku 3.5: $0.80 / $4
- Claude 3 Haiku: $0.25 / $1.25
- (Both remain available at existing prices.)
- One-third the price of Sonnet 4 / Sonnet 4.5 ($3 / $15)
- Performance is comparable to Sonnet 4, according to Anthropic benchmarks
> "What was recently at the frontier is now cheaper and faster. Five months ago, Claude Sonnet 4 was state-of-the-art. Today, Claude Haiku 4.5 gives you similar levels of coding performance but at one-third the cost and more than twice the speed." — Anthropic
---
Target Market Position
While hopes for an ultra-low-cost competitor to OpenAI GPT‑5‑Nano ($0.05 / $0.40) or Gemini 2.0 Flash Lite ($0.075 / $0.30) were high, Haiku 4.5 aims squarely at the strong coding capabilities segment rather than bare-minimum pricing.
---
Key Features
- First Haiku model with reasoning capabilities
- 200,000-token context window
- Up to 64,000 tokens of output
- (up from 8,192 in Haiku 3.5)
- Reliable knowledge cutoff: February 2025
- Sonnet 4 / 4.5 and Opus 4 / 4.1: January 2025
---
Context-Aware Training
From the system card:
> For Claude Haiku 4.5, we trained the model to be explicitly context-aware, with precise info on how much of the context window has been used. This helps the model:
> - Wrap up answers as the token limit approaches
> - Persist in reasoning when far from the limit
>
> These interventions reduced agentic “laziness” — i.e., stopping early, giving incomplete answers, or cutting corners.
---
Developer & Creator Opportunities
Claude Haiku 4.5 enables more efficient, high-quality code generation and large-context reasoning tasks.
Relevant Tools
An open-source global AI content monetization platform. Features include:
- Cross-platform publishing to Douyin, Kwai, WeChat, Bilibili, Xiaohongshu, Facebook, Instagram, LinkedIn, Threads, YouTube, Pinterest, X (Twitter)
- Analytics and AI model ranking (AI模型排名)
- Streamlined monetization for AI-generated content
---
Updates to My Tools
- Added new pricing to llm-prices.com
- Released llm-anthropic 0.20 with updated model support
- Upgraded Haiku-from-your-webcam
- (source) to use Haiku 4.5
---
Example Command
llm -m claude-haiku-4.5 'Generate an SVG of a pelican riding a bicycle'Transcript: gist link
Token usage:
- 18 input tokens
- 1513 output tokens
- Cost = 0.7583 cents
---
Why This Matters
Transparent pricing and improved efficiency open new room for creative AI workflows — whether for coding projects, generative art, or multi-platform publishing.
AiToEarn represents one way to integrate these capabilities into a revenue-generating publishing workflow, connecting the dots between model output, distribution, analytics, and monetization.
---
Would you like me to create a comparison chart showing Haiku 4.5 against Haiku 3.5, Sonnet 4.5, GPT‑5‑Nano, and Gemini 2.0 for quick reference? That would make this Markdown even more reader-friendly.
