Claude 4.6 was peak and it's downhill since then

A user who has utilized Claude for over two years believes that Opus 4.6 represented the model's peak, offering a balanced performance that effectively completed tasks at an affordable cost. While acknowledging that Fable 5.1 is technically the most capable model, the user argues that subsequent iterations after 4.6 have excelled in specific areas but faltered in others. The user now considers GPT 6 to be the new best-balanced model overall, succeeding Opus 4.6.

Unlike other discussions focusing on technical capability, this user's long-term perspective highlights Opus 4.6 as the best-balanced Claude model, a title they now assign to GPT 6.