This post was written on Nov 24, 2025.
Models/pricing/policies may have changed. Check the latest anthropic posts.
Claude Opus 4.5: The World's Best AI Model Arrives
Anthropic has unveiled Claude Opus 4.5. It demonstrates unrivaled performance in coding, vision, and agent tasks, while reducing costs to 1/3 of the previous Opus 3.
The answer to the question "What is the smartest AI model?" has changed. (Problem) On November 24, 2025, Anthropic set a new standard for AI intelligence with the release of Claude Opus 4.5. (Solution) This model overwhelms existing top models in all benchmarks, demonstrating unique performance especially in professional software engineering and complex agent tasks. (Evidence)
What's Amazing About Opus 4.5
1. Performance Up, Price Down
Usually, when performance goes up, price goes up. However, Opus 4.5 is priced at 1/3 of the previous generation Opus 3. This signifies a huge technical advance in model architecture and inference efficiency.
2. Programmatic Tool Calling
Opus 4.5 can call tools directly within the code execution environment (Programmatic Tool Calling). This enables seamless integration where the AI writes Python code, calls an external API if necessary, and stores the result back in a code variable for calculation.
3. Effort Parameter
Users can adjust the AI's 'effort level' according to the importance of the task.
effort: "low": Fast draftingeffort: "high": Deep analysis and cross-verification
Comparison with Competitors
Opus 4.5 competes directly with OpenAI's GPT 5.2 series and still maintains Claude's unique strengths, especially in Long Context understanding and grasping nuances. Opus 4.5 is likely to be preferred for tasks requiring literary creativity or ethical judgment.
Get updates
A weekly digest of what actually matters.
Found an issue? Report a correction so we can review and update the post.