What happened
Anthropic on Wednesday launched Claude Haiku 5.5, the third model in its Claude 5.5 family in the past month, completing the lineup ahead of its planned IPO. Haiku 5.5 is the speed-and-cost workhorse of the family: built for classification, summarization, and extraction — think live customer support, voice agents, and in-app assistants — at 75% lower cost than its predecessor, Haiku 4.5.
The details
Haiku 5.5 is priced at $0.10 per million input tokens and $0.50 per million output tokens for prompts under 100,000 tokens, rising to $0.50 and $2.50 respectively for longer prompts. That is deep in “cheapest frontier-tier option” territory — the kind of pricing that makes voice agents and always-on support bots economically viable at real scale.
Notably, Anthropic says Haiku 5.5 is the first Haiku model with built-in safeguards for a narrow set of high-risk cybersecurity requests, adding that most everyday tasks will not be impacted. The 5.5 family now spans three tiers — Opus for the hardest reasoning work, Sonnet for coding and agents, Haiku for high-volume speed tasks — released in rapid succession over the past few weeks.
Why it matters
This is Anthropic playing the lineup game that investors watch before an IPO: a full, legible product stack with a clear price ladder. Haiku 5.5 matters most to builders — anyone running a support bot, voice agent, or real-time classifier gets frontier-adjacent quality at a fraction of the cost, while the safeguard note is a quiet nod to regulators and enterprise buyers asking what “responsible speed” looks like. Bottom line: if you are paying for Haiku 4.5 or renting bigger models for grunt work, Haiku 5.5 just cut your bill — just check the 100,000-token cutoff before migrating long conversations.
Frequently asked questions
What is Claude Haiku 5.5?
Anthropic’s newest fast, low-cost AI model, released October 7, 2026. It is the third in the Claude 5.5 family, designed for high-volume tasks like classification, summarization, live support, and voice agents.
How much does Haiku 5.5 cost?
$0.10 per million input tokens and $0.50 per million output tokens for prompts under 100,000 tokens — 75% less than Haiku 4.5. Longer prompts rise to $0.50/$2.50 per million.
Should I use Haiku 5.5 instead of Sonnet 5.5?
For speed-critical, high-volume work (support, voice agents, extraction), Haiku 5.5 is the right pick. For coding, complex reasoning, and agentic work, Sonnet 5.5 remains the stronger model.
Is Anthropic going public?
Reuters describes the launch as coming ahead of a planned IPO; earlier reports have pointed to marketing as early as November 2026, but no date has been officially confirmed.
Sources: Reuters