You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I have verified this feature I'm about to request hasn't been suggested before.
Describe the enhancement you want to request
Add collapsible UI for reasoning summary parts, matching the existing "Explored" pattern. Shows "Reasoning" with shimmer animation while the model is thinking, then transitions to "Reasoning Summary" as a collapsible section (open by default) when complete.
Reacted by Byeonghoon Yoo, Chino Chang, Houssam Elbadissi, Diogo Sousa, Ernesto Baschny, BlankDev117, Sté Thunus, Yihao Gao, seajoyhill and Father Gucci
+1 to having collapsed thinking, as currently when showing thinking is enabled, the TUI conversation feels rather noisy, and hard to visually ingest.... compared to Claude Code and Codex in comparison.
Reacted by RJ, Houssam Elbadissi, streeyt and Father Gucci
Can affirm the same. I almost always want to see the reasoning as the model does it (particularly for open-weight models which show the actual full reasoning), so sometimes it's easier to stop it and steer if it goes too deep the wrong road or gets fixated on something that's not consequential to my request (e.g. random errors or warning that aren't relevant to the task).
I think the best implementation would be to automatically expand the "current" reasoning block, and automatically collapse it once it moves to the next reasoning block (or finishes the actual final response), with the ability to expand any past reasoning block.
Reacted by Will LE, grumpypotato-bambu, Sté Thunus, 0-SamboNZ-0 and Father Gucci
Activity
github-actions commentedon Feb 26, 2026
This issue might be a duplicate of existing issues. Please check:
If your request is distinct, please clarify how it differs from the above.
epicwhale commentedon Apr 24, 2026
+1 to having collapsed thinking, as currently when showing thinking is enabled, the TUI conversation feels rather noisy, and hard to visually ingest.... compared to Claude Code and Codex in comparison.
yourjhay commentedon May 1, 2026
+1 This is very useful feature if ever. Making opencode TUI more clean.
skaldebane commentedon May 10, 2026
Can affirm the same. I almost always want to see the reasoning as the model does it (particularly for open-weight models which show the actual full reasoning), so sometimes it's easier to stop it and steer if it goes too deep the wrong road or gets fixated on something that's not consequential to my request (e.g. random errors or warning that aren't relevant to the task).
I think the best implementation would be to automatically expand the "current" reasoning block, and automatically collapse it once it moves to the next reasoning block (or finishes the actual final response), with the ability to expand any past reasoning block.
github-actions commentedon Jul 26, 2026
To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.
fidodido48 commentedon Aug 14, 2026
How about expanding just code blocks (w/o reasoning/thinking)? 'Code concealment' doesnt work ...