Anthropic has introduced its latest AI marvel, the Claude 3.5 Haiku, which promises to surpass its predecessors in performance on many fronts. But with all its advancements, it brings a heftier price tag, and is missing one significant capability: image analysis.
Last month, Anthropic revealed Claude 3.5 Haiku, a model set to challenge or even exceed the benchmarks of their previous top-tier AI, the Claude 3 Opus. Accessible via Anthropic’s API and available on platforms like AWS Bedrock, Claude 3.5 Haiku is designed for tasks such as coding suggestions, data extraction, labeling, and content moderation.
Originally, Anthropic suggested that the cost of Claude 3.5 Haiku would mirror that of its predecessor. However, after discovering that the new model outshines Claude 3 Opus by a significant margin, Anthropic adjusted the pricing accordingly. In a statement on social media platform X, they explained this decision, highlighting the enhanced intelligence of their latest model.
The pricing shift is notable. The Claude 3.5 Haiku now starts at $1 per million input tokens, approximately equating to 750,000 words, and $5 per million for output tokens. This is a stark increase from the 25 cents per million input tokens and $1.25 per million output tokens for Claude 3 Haiku.
On the downside, Claude 3.5 Haiku doesn’t come with image analysis features initially, unlike the Claude 3 Haiku. Alex Albert from Anthropic confirmed that the earlier model remains available for those who prioritize cost-efficiency and require image processing capabilities.
Despite this, Claude 3.5 Haiku can generate longer text outputs and has access to more recent information, broadening its application scope. Nonetheless, the decision to elevate the model’s pricing raises eyebrows and questions about Anthropic’s future pricing strategies. As they move forward, industry watchers will be keen to see how these pricing changes impact user adoption and the broader AI market landscape.
