Anthropic Releases Claude 3.5 Sonnet With Top-Tier Speed and Vision
Anthropic unveils Claude 3.5 Sonnet, a powerful new AI model that outperforms its predecessors in reasoning and coding while operating at twice the speed. The model is available across multiple platforms and excels at complex visual reasoning tasks.
Anthropic launches Claude 3.5 Sonnet, the first model in its new Claude 3.5 family, setting a new standard for AI intelligence. The model outperforms competitor models and the older Claude 3 Opus across numerous evaluations while maintaining the speed and cost efficiency of the mid-tier Claude 3 Sonnet. Users access this upgraded intelligence for free on Claude.ai and the iOS app, with higher rate limits available for Pro and Team subscribers.
The new model operates at twice the speed of Claude 3 Opus and achieves state-of-the-art results in graduate-level reasoning, undergraduate-level knowledge, and coding proficiency. In internal agentic coding tests, Claude 3.5 Sonnet successfully resolves 64% of problems compared to Opus's 38%, allowing it to independently write, edit, and execute code to fix bugs or add features. This combination of rapid performance and sophisticated troubleshooting makes the model highly effective for multi-step workflows and updating legacy codebases.
Claude 3.5 Sonnet also introduces major upgrades in visual capabilities, surpassing Claude 3 Opus on standard vision benchmarks. The model demonstrates exceptional skill at visual reasoning tasks like interpreting charts and graphs, and it accurately transcribes text from imperfect images. These vision improvements provide significant value for industries like retail, logistics, and finance, where businesses extract critical insights from complex graphics and documents.