✳ the wire · analysis

Claude Haiku 5.5 is out at $0.10 / $0.50, a tenth of Haiku 4.5's price

Claude Haiku 5.5confirmedby ArtificialWatch

An email when every new Anthropic model answers on a public API, about 15 minutes after members, plus every frontier launch · no card

ArtificialWatch title card · verified against a primary source

Anthropic released Claude Haiku 5.5 on October 7 as claude-haiku-5-5, on the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry and in the Claude apps. It is priced in two tiers by prompt length: $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens, and $0.50 / $2.50 above that. Our launch sweep fired on the new id at 17:52 UTC.

THE PRICE

  • Prompts up to 100K tokens: $0.10 input, $0.50 output, $0.01 cache reads, $0.125 cache writes (5-minute)
  • Prompts over 100K tokens: $0.50 input, $2.50 output, $0.05 cache reads
  • Batch API: half of either tier
  • Haiku 4.5 costs $1 / $5, so short prompts now cost a tenth as much

SPECS

A 1M-token context window, 128K max output, adaptive thinking at medium effort by default, and a June 2026 knowledge cutoff. Anthropic commits to keeping it available until at least October 7, 2027.

ANTHROPIC'S OWN FIGURES

Anthropic's launch table compares it with Haiku 4.5, OpenAI's GPT-6 Luna (which costs the same $0.10 / $0.50) and Sonnet 5.5:

  • OSWorld 2.1 (computer use): 72.4%, against 15.7% for Haiku 4.5 and 48.9% for GPT-6 Luna
  • Terminal-Bench 4.0: 39.2%, against 0.0% and 16.4%
  • GDPval-AA v2.1: 1,620, against 735 and 1,437
  • FrontierCode 1.1: 46.4%, against 42.4% for GPT-6 Luna
  • Humanity's Last Exam: 45.9% without tools and 57.4% with them, against 10.2% and 18.7% for Haiku 4.5

Sonnet 5.5 stays ahead on every row, for example 70.6% on Terminal-Bench 4.0.

ALSO TODAY

Anthropic halved Sonnet 5.5's cache-read price to $0.10 per million tokens, from $0.20, and says that makes most long-running work about 20% cheaper.

FIRST INDEPENDENT SCORE

Artificial Analysis scores it 43 on its Intelligence Index at max effort, second of 179 models in its price tier, where the median is 13. It used 440M tokens to run the index, against a median of 120M, which Artificial Analysis rates as very verbose.

WHAT WE DO NOT KNOW

How it does on the other independent boards we track, and its output speed, which Artificial Analysis has not measured yet. OpenRouter listed it at 18:31 UTC at the same two-tier prices.

Source: Anthropic (launch page, models overview and pricing docs, read directly) · Artificial Analysis ↗ · Claude Haiku 5.5 tracker · the bench index
Data: LLM knowledge cutoffs · LLM API pricing · LLM context windows

Cite this page

Writing about this story? Copy a link or a citation.

<a href="https://artificialwatch.com/wire/claude-haiku-5-5-launch">Claude Haiku 5.5 is out at $0.10 / $0.50, a tenth of Haiku 4.5's price</a>
ArtificialWatch, "Claude Haiku 5.5 is out at $0.10 / $0.50, a tenth of Haiku 4.5's price", https://artificialwatch.com/wire/claude-haiku-5-5-launch
sweeping every 60 seconds

Hear when every new Anthropic model drops — not when we write it up.

Claude Opus 5 went live at 16:51 UTC. Members had the alert at 16:52. Free gets a Chrome push and an email about 15 minutes after members.

  • Free forever
  • No card
  • Unsubscribe in one click

Want it the minute it answers? Try Signal 7 days free →

← back to the wire · every story, by month