Anthropic: Claude 3.7 Sonnet
Released Recently|128K context|$3.00/M input tokens|$15.00/M output tokens
Claude 3.7 Sonnet is Anthropic's first hybrid reasoning model that can produce near-instant responses or extended, step-by-step thinking. It excels at complex tasks in math, physics, and coding, with 45% fewer unnecessary refusals. The model can control thinking time and budget through API, making it ideal for both quick interactions and deep problem-solving.
Speed
Intelligence
Price (1M Tokens)$3.00
Inputs
TextImage
Outputs
Text
Pricing
Input
$3.00
Cached input
$0.75
Output
$15
Claude 3.7 Sonnet
$3.00
GPT-4o
$2.50
GPT-4.1
$2.00
Use Case
Complex reasoning
Handle the most demanding tasks requiring frontier intelligence
Mathematical proofs
Advanced AI capability for specialized tasks
Physics problems
Advanced AI capability for specialized tasks
Coding tasks
Advanced coding with state-of-the-art performance (72.5% SWE-bench)
Front-end development
Advanced AI capability for specialized tasks
Quick interactions
Advanced AI capability for specialized tasks
Features
Hybrid Reasoning
Near-instant or extended step-by-step thinking
Thinking Budget Control
Control thinking time via API
Improved Coding
Enhanced coding and front-end web development
Reduced Refusals
45% fewer unnecessary refusals
Tool Use
Interact with external tools and services to enhance capabilities.
Computer Use
Support for computer use functionality.