What Is Claude Haiku 5.5? Features, Context Window & Pricing
Claude Haiku 5.5 is Anthropic's newest Haiku model: a 1M-token context window, 128K output, adaptive thinking with effort levels and $0.10 / $0.50 API pricing. Here's what it can do and how to try it free.
Haiku55 Team
Claude Haiku 5.5 is the newest Haiku-class model from Anthropic — the fast, low-cost tier of the Claude family. It succeeds Claude Haiku 4.5 and is built for high-volume, latency-sensitive work: chat, classification, routing, extraction and sub-agent tasks inside larger AI agents.
You can try Haiku 5.5 free on Haiku55 — no sign-up needed.
Claude Haiku 5.5 at a glance
| Spec | Claude Haiku 5.5 |
|---|---|
| Model ID | claude-haiku-5-5 |
| Context window | 1,000,000 tokens |
| Max output | 128,000 tokens |
| Thinking | Adaptive, on by default |
| Effort levels | low, medium (default), high, xhigh, max |
| API price (prompts ≤ 100K tokens) | $0.10 input / $0.50 output per 1M tokens |
| API price (prompts > 100K tokens) | $0.50 input / $2.50 output per 1M tokens |
What's new in Haiku 5.5
A 1M-token context window
Haiku 4.5 topped out at 200K tokens. Haiku 5.5 accepts up to one million tokens of input, so a long report, a bundle of contracts or a mid-sized codebase can fit into a single conversation. The output limit doubles too, from 64K to 128K tokens.
Adaptive thinking with effort levels
Haiku 5.5 thinks before it answers by default — and decides for itself how much thinking a request needs. Instead of a fixed thinking budget (the Haiku 4.5 approach), you choose an effort level:
- low — the fastest and cheapest; ideal for chat and simple requests
- medium — the default and a good start for most work
- high — knowledge work, longer agent tasks and strict instruction following
- xhigh / max — the most thorough, when quality matters more than speed
On Haiku55 these map to the Fast, Smart and Deep modes in the chat box.
Better instruction following and agent work
According to Anthropic, Haiku 5.5 is substantially better than Haiku 4.5 at following instructions and at running as a sub-agent, across coding and knowledge work. It supports server-side tools such as web search and web fetch — the reason the Haiku55 agent can look things up and cite its sources.
Lower prices
Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100K tokens, versus $1 / $5 for Haiku 4.5. One caveat: Haiku 5.5 uses a newer tokenizer, so the same text counts as roughly 30% more tokens.
What is Haiku 5.5 good for?
- Everyday chat and Q&A — fast answers in most languages
- Research with sources — pair it with web search for current, cited information
- Summarizing and translating long documents and web pages
- Extraction and classification at scale
- Coding help — explanations, snippets and quick reviews
- Sub-agents inside larger AI systems, where speed and cost matter most
How to use Claude Haiku 5.5 for free
- Open Haiku55 — the chat is right on the homepage.
- Ask anything. Guests get free messages every day.
- Sign in to unlock web search, the Smart and Deep modes, saved history and free daily credits.
Developers can also call the model through the Claude API with the model ID claude-haiku-5-5.
Haiku55 is an independent service and is not affiliated with Anthropic. Claude and Haiku are trademarks of Anthropic, PBC.