The market for artificial intelligence (AI) models is especially hot at the moment, and Anthropic is taking note.
On Thursday, several Claude users reported via social media that the company had just made its latest model in that family, 3.5 Haiku, generally available to all users of the chatbot.
Also: How does Claude work? Anthropic reveals its secrets
Originally announced in October, 3.5 Haiku is, Anthropic says, the company’s fastest model, “ideal for software teams looking to streamline their coding process and boost productivity.” Other suggested use cases include data extraction, content moderation, personalization, and building chatbots that require scalable engagement.
Although the release has been confirmed by several outlets and users, Claude itself seemed a bit confused about whether it was running the latest version of Haiku. When I asked, Claude stated that its sibling model, Claude 3.5 Sonnet, is the most up-to-date.
In terms of benchmarks, 3.5 Haiku is comparable to Claude 3.5 Sonnet, but currently ranks below both Claude 3 Opus and 3.5 Sonnet in the Chatbot Arena. Overall, however, Haiku is reportedly cheaper and higher-quality than average, and outperforms several popular models, especially in coding. Gemini 1.5, GPT-4o, 3.5 Sonnet, and o1 are more expensive, and 3.5 Haiku is faster than Gemini 1.5, Llama 3.1 405B, and 3.5 Sonnet.
The model also can be used with Artifacts, a sidebar feature launched in June that “can create and display the fully formatted results of your request in real-time and side-by-side with your conversation,” as ZDNET’s Lance Whitney explains.
<!–>
Anthropic has yet to officially announce the release. As with many free tiers for AI chatbots, users will likely experience daily messaging limits, but can upgrade to the Claude Pro plan for $20 a month (or $18 per month billed annually) if they want more usage.
