Launched in June 2024,
Claude 3.5 Sonnet is a significant leap in AI technology, developed by Anthropic. It raises the bar for intelligence across various benchmarks such as
graduate-level reasoning (GPQA),
undergraduate-level knowledge (MMLU), and
coding proficiency (HumanEval). This model is touted to be
twice as fast as its predecessor, Claude 3 Opus, and offers a
200K token context window, which is larger than GPT-4o's 128K token capacity. The model's pricing is also competitive at
$3 per million input tokens and
$15 per million output tokens. You can learn more straight from the
source.
These features set Claude 3.5 apart from typical models, enabling it to excel in multi-step workflows and context-sensitive tasks, such as customer support.
For a detailed breakdown of performance, the comparison was documented in the
Medium blog post with score tallies favoring GPT considerably in various aspects.