Jan 16
Loading video player...
In this video, I share my honest thoughts on Opus 4.5 and why I believe it is finally the true successor to Sonnet 3.5 for coding. We discuss the massive price reduction that makes it viable, its incredible ability to handle long-context workflows, and how Anthropic is winning by treating their models like actual products. -- Key Takeaways: šø Opus 4.5 brings a massive price reduction (approx. 3x cheaper), making it accessible for heavy coding workflows. š This model is the true recreation of the 3.5 Sonnet experience, but significantly more powerful for end-to-end tasks. š ļø Unlike previous models, Opus 4.5 religiously checks for linting and build errors before finishing a task. š While benchmarks might show small gains, the "real feel" performance is 2x better than Sonnet 4.5. š¢ Anthropic is succeeding by listening to developer feedback and fixing bugs, unlike the "compute-first" strategy of other labs. š¤ The model can handle full project lifecycles, performing dozens of tool calls and working for long durations autonomously. š This is the "great unlock" we were expecting from Gemini 3, delivering consistent, high-quality code generation.