Anthropic: Claude Opus 4.1

The Anthropic model is designed for tasks that require complex reasoning and handling of multiple input types. It supports three input modalities: image, text, and file inputs, with a maximum context length of 200000 tokens. The model also offers tools and reasoning capabilities, but does not provide structured output. Those considering this model may want to shortlist it due to its broad range of supported tasks and modalities. However, its high price point should be taken into account: $15 per million input tokens and $75 per million output tokens. Its blended benchmark score is 29.8 across one independent benchmark, which is a decent standing but not exceptional considering the price. This model may be worth exploring for those who require specific capabilities or can justify the cost.