Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
Claude 3.5 Sonnet surpasses the previous Claude 3 Opus with lower running costs and enhanced visual capabilities
Anthropic, the Amazon-backed challenger to OpenAI, has rolled out the first major upgrade to its flagship Claude 3 models.
The new Claude 3.5 Sonnet is a mid-tier model, designed to improve on Claude 3’s performance capabilities.
Upon unveiling the new model, Anthropic said it “raises the industry bar for intelligence” and outperforms rival models, including OpenAI’s new GPT-4o and Gemini 1.5 Pro from Google.
Claude 3.5 Sonnet is also designed to be less expensive to run than previous Anthropic models, costing just $3 per million input tokens and $15 per million output tokens.
The model has a token context window of 200,000 tokens, meaning maxing out the amount of text it can handle in one prompt would cost just $0.60 for the input.
Credit: Anthropic
The new Claude 3.5 Sonnet model is available on the Claude web application and iOS app.
Claude Pro and Team subscribers get access to the new model at higher rate limits than free users.
Enterprises using Anthropic’s API or third-party platforms like Amazon Bedrock and Google Cloud’s Vertex AI can also access the model.
Amazon, which has invested $4 billion in the OpenAI rival, said Claude 3.5 Sonnet is “the ideal model for complex tasks such as context-sensitive customer support, orchestrating multi-step workflows, streamlining code translations and creating revenue-generating user-facing applications.”
Claude 3.5 Sonnet is the first of what will be an entire line of upgraded Claude 3 models.
New 3.5 versions of the small model Haiku and large model Opus will launch “later this year.”
The latest version of Sonnet boasts improved writing abilities, with the model able to generate content with improved nuance and tones.
The model generates content twice as fast as the largest Claude 3 model, Opus.
It also has improved vision capabilities and can interpret charts and graphs and transcribe text in imperfect images.
In terms of performance, it achieved high scores on evaluation benchmarks including MMLU for general knowledge and coding test HumanEval.
The model outshone the prior Claude 3 Opus and even outscored GPT-4o and Gemini 1.5 Pro on model evaluations including MathVista, AI2D and ANLS Score.
“When instructed and provided with the relevant tools, Claude 3.5 Sonnet can independently write, edit, and execute code with sophisticated reasoning and troubleshooting capabilities,” Anthropic said. “It handles code translations with ease, making it particularly effective for updating legacy applications and migrating codebases.”
Credit: Anthropic
In addition to the upgraded models, Anthropic announced it’s working on expanding Claude to new modalities. It also plans to add new features to support more use cases for businesses, including integrations with enterprise applications.
One feature Anthropic is working on is memory — which would let Claude recall user preferences and interaction history. OpenAI has already been working on this for ChatGPT, with Anthropic saying it would make user experiences on Claude “even more personalized and efficient.”
“We look forward to seeing what you build, create and discover with Claude,” the startup said.
Read more about:
ChatGPT / Generative AIYou May Also Like