Expert & Genius intelligence levels now available
by Andrew Lee
Today, we’re introducing our brand new Expert and Genius intelligence levels to give you the most power possible when automating your complex business processes.
Expert uses Claude Opus 4.5 for high-capability task execution. It’s ideal for demanding workflows—complex research, multi-step automations, and anything that needs real reasoning depth.
Genius takes it further. It pairs Opus 4.5 with extended thinking and enhanced context, giving the model more room to reason through the hardest problems.
Expert is available on the Premier plan and Genius on the Max plan.
Anthropic’s most powerful model yet
In Anthropic’s testing, Opus 4.5 outperformed every human candidate on their notoriously difficult engineering take-home exam. Partners report it handles complex multi-step tasks with fewer dead-ends and “just gets it” in a way previous models didn’t.
For Tasklet users, this translates into agents that plan better, reason more effectively through complex problems, recover from errors more gracefully, and handle ambiguity without hand-holding. Tasks that used to require back-and-forth now complete on the first try. Whether you’re navigating complex compliance workflows or synthesizing research across dozens of sources, Opus brings the reasoning depth to match.
The right tool for the job
Not every task needs Genius-level intelligence. Higher intelligence levels consume more quota (Expert uses roughly 5x and Genius ~10x). A simple daily briefing can run at Basic. A multi-source research project deserves Expert or Genius. Tasklet gives you the flexibility to match capability to complexity—and manage your quota accordingly.
You can pick intelligence levels per-agent using the selector next to the message input. If you want every new agent to start at a specific level, set your default in Settings → General.
Have feedback? Drop me a note at andrew@tasklet.ai.