I was missing Claude.

After switching from Claude to gpt 5.6 sol and then Astra for about a month and a half, a user expressed relief and satisfaction upon returning to Claude. Despite finding OpenAI productive, the user felt something was missing, describing Claude as having a "gentle touch" and being more polished and refined. They noted that Claude "gets you more and got shits done," suggesting a preference for its user experience and effectiveness.

This user's direct comparison of Claude with gpt 5.6 sol and Astra highlights Claude's perceived superior "gentle touch" and refinement, unlike other models.