Anthropic Launches Claude Haiku 4.5 And Unveils Agent Skills To Enhance Tasks

As September ended, Anthropic unveiled Claude Sonnet 4.5, a new frontier model that the artificial intelligence (AI) startup said delivers better abilities and even dubbed it “the best coding model in the world.”

The announcement came just three days after it revealed plans to grow its global presence, triple its international workforce, and scale up its applied AI team 5x.

Now, Anthropic has introduced another model—Claude Haiku 4.5—its latest small model, which is now available to all users.
 

Claude Haiku 4.5

According to the company, the new model offers users similar levels of coding performance as Claude Sonnet 4—which was a state-of-the-art model five months ago—but at one-third the cost and more than twice the speed. It even outshines Claude Sonnet 4 at certain tasks, such as using computers.

Claude Haiku 4.5’s high intelligence and remarkable speed will be appreciated by users who leverage AI for real-time, low-latency tasks such as chat assistants, customer service agents, or pair programming. Claude Code users will note that the new model can make the coding experience more responsive, spanning multiple-agent projects, rapid prototyping, and more.

It offers users near-frontier performance with much greater cost-efficiency and brings new ways of using multiple models together. “For example, Sonnet 4.5 can break down a complex problem into multi-step plans, then orchestrate a team of multiple Haiku 4.5s to complete subtasks in parallel,” says Anthropic.

Claude Haiku 4.5 is priced at $1/$5 per million input and output tokens and is available on Claude Code and Anthropic’s apps—developers can access it through Anthropic's API, Amazon Bedrock, and Google Cloud’s Vertex AI.

“Claude Haiku 4.5 hit a sweet spot we didn't think was possible: near-frontier coding quality with blazing speed and cost efficiency,” said Guy Gur-Ari, the co-founder of Augment Code. “In Augment's agentic coding evaluation, it achieves 90% of Sonnet 4.5's performance, matching much larger models. We're excited to offer it to our users.”

 

Agent Skills

Anthropic also announced a new feature called “Skill,” which can help Claude improve how it performs specific, specialized tasks, such as working with Excel or following a company’s brand guidelines.

Skills are folders that include instructions, scripts, and resources that Claude can load when needed. These skills will only be accessed by Claude when they’re relevant to the task being performed.

When working on tasks, Claude automatically scans and identifies available skills to find relevant matches and coordinates their use. It only loads what's needed, when it's needed, keeping Claude fast while accessing specialized expertise.

Skills is already being used in Claude apps, helping Claude create files such as spreadsheets and presentations. Now, users can build their own skills to be used across Claude apps, Claude Code, and the API, as Skills use the same format everywhere.

 

Salesforce Partnership Expansion

Anthropic and Salesforce announced an expanded partnership that will see Claude become a preferred model for Salesforce’s Agentforce platform.

The move will bring Salesforce financial services, healthcare, cybersecurity, and life sciences customers a trusted AI that keeps sensitive data secure.

Salesforce will also deploy Claude Code across its global engineering organization to augment developer workflows and increase productivity, while Anthropic broadens its use of Slack by infusing Claude directly, to summarize conversation threads, surface key insights, and guide the internal sales team.

The expanded partnership comes with three new initiatives:
 

• Providing Claude as a preferred foundational model in Agentforce via Amazon Bedrock.

• Building tailored, industry-specific AI solutions, beginning with financial services.

• Deepening integration between Claude and Slack, including bringing Claude access to Slack channels, messages, and files to summarize conversations, extract key decisions, and draft updates.

“Regulated industries need frontier AI capabilities, but they also need the appropriate safeguards before they can deploy in sensitive systems. We've built Claude to deliver both: the performance and the safeguards,” said Dario Amodei, CEO and Co-Founder of Anthropic.

“By bringing Salesforce directly into Claude—and Claude’s intelligence into Salesforce and Slack—we’re giving every company the power to work in entirely new ways,” said Marc Benioff, Chair and CEO of Salesforce. “Together, we’re making trusted, agentic AI real for every industry—combining Anthropic’s world-class models with the trust, reliability, accuracy and scale of Agentforce 360, helping customers achieve new levels of productivity, innovation, and growth.”

Anthropic also said Claude now integrates with Microsoft 365 and offers enterprise search across connected tools, which is available for all Claude Team and Enterprise plan customers.

In Microsoft 365, Claude helps users search and analyze documents across sites and libraries in Microsoft SharePoint and OneDrive, Access email threads, analyze communications patterns, and extract insights in Microsoft Outlook, and search through chat conversations, channel discussions, and meeting summaries to offer key insights in Microsoft Teams.

Anthropic’s Claude AI models are popular globally, especially in the enterprise AI sector, where it holds the top market share.

Do you think these moves will help Anthropic chip away at OpenAI’s global market share and help it capture a commanding position in the AI sector?

Let us know in the comments below!