Urgent.News

What's breaking now, across hundreds of outlets.

AI

Anthropic upgrades Claude with new Haiku 5.5 model, details here

Anthropic has released its latest AI model with Claude Haiku 5.5. The new version arrives about a year after the previous Haiku model upgrade. more…

Anthropic upgrades Claude with new Haiku 5.5 model, details here

Anthropic recently unveiled its latest AI model, Claude Haiku 5.5, marking a year since their previous upgrade. Claude Haiku is the smallest and cheapest AI model offered by Anthropic, alongside its medium-sized Sonnet and large-capacity Opus models. The platform also features a more advanced and expensive Fable/Mythos model for users seeking higher capabilities.

With this release, Claude Haiku has been upgraded from version 4.5 to 5.5, the third Haiku model upgrade within a month, following Opus and Sonnet improvements. Anthropic explains that Claude Haiku 5.5 is tailored for high-volume, cost-sensitive tasks such as summaries, compactions, database queries, and classification requests. The model works seamlessly with Opus 5.5 and Sonnet 5.5 as a subagent for coding work and excels in speed-sensitive tasks like live customer support and browser usage.

Additionally, Anthropic highlights that Haiku 5.5 is around 75% less expensive to run compared to its predecessor. The company also announced price reductions for Claude Sonnet 5.5, halving the cost of cache reads, making it approximately 20% cheaper for most agentic work. Furthermore, Anthropic introduced a new monthly API credit for Claude Max and Team subscribers, aiming to facilitate the development of new agents and applications on the Claude Platform.

In comparison to Anthropic's other models, Haiku 5.5 is positioned as a low-cost alternative to OpenAI's GPT-5.5 Luna, another budget-friendly AI model. Users can find more information about Claude Haiku 5.5 on Anthropic's website.

Rewritten by urgent.news; may contain errors — check the original.

This story

This is one outlet's version. Read the fullest account.

Every outlet that reported this, and when →

Read the original at 9to5mac.com →

Report a problem with this story

More in AI

Five Ways My AI Agents Went Wrong (and What I Changed)

I run a small software company by myself. Claude Code agents play the engineers, QA, marketing and so on. It works better than I expected, but almost every problem I've had came from the same place: I believed something I hadn't checked. An agent's "done," a file the agents read, a doc that looked current.

More from Wednesday 7 October →