Activity

1. opencode-agent commented on Oct 21, 2026

   

   opencode-agentBot

   on Oct 21, 2026 – with OpenCode Agent

   Contributor

   More actions

   Thanks for the report, @zzz-creator!

   In 2.1.2, the TUI collapses thinking by default. The model's reasoning shows as a single dim line, "Thinking…" while it runs and "Thought" (sometimes "Thought · N steps") when it's done. Click that line to open it. To always show it in full, run Expand thinking from the command palette, or set Settings → Session → Thinking to show (this is session.thinking: "show" in your TUI config).

   Also, the model's own answer ("the thinking process isn't shown…") doesn't tell us much. The model can't see how OpenCode displays its output.

   To work out whether this is that setting or a bug, could you tell us:

   1. Do you see any "Thinking…" / "Thought" lines between the file reads, searches and commands? Or is there nothing at all?
   2. After switching thinking to show, does anything change?
   3. Does any of the reasoning end up inside the answer text instead, sometimes with a literal </think>?
   4. Does the same thing happen with another model that supports thinking?

   You can reply here, or edit the issue. If you edit it, please also leave a comment saying you've updated it, so we know when you're done.

   If no Thought line appears at all, these existing issues about Big Pickle may already cover it (most likely first):

   • reasoning: reasoning text missing entirely despite reasoning tokens (thinking never shown, no text leak) #54886: Big Pickle uses reasoning tokens, but no thinking block is saved or shown (also reported on Windows)
   • reasoning: opencode zen (big-pickle) inline </think> leaks into text content instead of a reasoning part #54599: Big Pickle's reasoning arrives inside the answer text instead of as a thinking block
   • provider: big-pickle leaks chain-of-thought into visible text when reasoning contains the literal close-think tag #55268: Big Pickle's reasoning spills into the visible reply when </think> appears inside it
2. opencode-agent

   added

   triaging

   on Oct 21, 2026
3. zzz-creator commented on Oct 21, 2026

   

   zzz-creator

   on Oct 21, 2026

   Author

   More actions

   Do you see any "Thinking…" / "Thought" lines between the file reads, searches and commands? Or is there nothing at all?
   Nothing.
   After switching thinking to show, does anything change?
   How do i do that?
   Does any of the reasoning end up inside the answer text instead, sometimes with a literal ?
   No
   Does the same thing happen with another model that supports thinking?
   Seems to be just big pickle.
4. zzz-creator commented on Oct 21, 2026

   

   zzz-creator

   on Oct 21, 2026

   Last edited by zzz-creator

   Author

   More actions

   @opencode-agent
   Do you see any "Thinking…" / "Thought" lines between the file reads, searches and commands? Or is there nothing at all?
   Nothing.
   After switching thinking to show, does anything change?
   How do i do that?
   Does any of the reasoning end up inside the answer text instead, sometimes with a literal ?
   No
   Does the same thing happen with another model that supports thinking?
   Seems to be just big pickle.
5. zzz-creator commented on Oct 21, 2026

   

   zzz-creator

   on Oct 21, 2026

   Last edited by zzz-creator

   Author

   More actions

   Thanks for the report, @zzz-creator!

   In 2.1.2, the TUI collapses thinking by default. The model's reasoning shows as a single dim line, "Thinking…" while it runs and "Thought" (sometimes "Thought · N steps") when it's done. Click that line to open it. To always show it in full, run Expand thinking from the command palette, or set Settings → Session → Thinking to show (this is session.thinking: "show" in your TUI config).

   Also, the model's own answer ("the thinking process isn't shown…") doesn't tell us much. The model can't see how OpenCode displays its output.

   To work out whether this is that setting or a bug, could you tell us:

   1. Do you see any "Thinking…" / "Thought" lines between the file reads, searches and commands? Or is there nothing at all?
   2. After switching thinking to show, does anything change?
   3. Does any of the reasoning end up inside the answer text instead, sometimes with a literal </think>?
   4. Does the same thing happen with another model that supports thinking?

   You can reply here, or edit the issue. If you edit it, please also leave a comment saying you've updated it, so we know when you're done.

   If no Thought line appears at all, these existing issues about Big Pickle may already cover it (most likely first):

   • reasoning: reasoning text missing entirely despite reasoning tokens (thinking never shown, no text leak) #54886: Big Pickle uses reasoning tokens, but no thinking block is saved or shown (also reported on Windows)
   • reasoning: opencode zen (big-pickle) inline leaks into text content instead of a reasoning part #54599: Big Pickle's reasoning arrives inside the answer text instead of as a thinking block
   • provider: big-pickle leaks chain-of-thought into visible text when reasoning contains the literal close-think tag #55268: Big Pickle's reasoning spills into the visible reply when </think> appears inside it

   Do you see any "Thinking…" / "Thought" lines between the file reads, searches and commands? Or is there nothing at all?
   Nothing.
   After switching thinking to show, does anything change?
   already on
   Does any of the reasoning end up inside the answer text instead, sometimes with a literal ?
   No
   Does the same thing happen with another model that supports thinking?
   Seems to be just big pickle.
6. sashalimarev commented on Oct 21, 2026

   

   sashalimarev

   on Oct 21, 2026

   More actions

   Also seeing this on 2.1.2, macOS/Ghostty, with OpenAI GPT-6.1 Sol. Thinking is set to Show, and reasoningSummary is set to auto. My JSON export contains nonempty reasoning text, but I don’t see it in the terminal. Different model and platform from the original report, so I’m not sure it’s the same cause.