Reasoning tokens not expanding in Codex UI - "Worked for X" shows time but no trace

[Rezhnn]

What happens

When I use opencodex with any proxy model (DeepSeek v4 flash free, Mimo v2.5 free, etc.), the Codex app shows "Worked for Xs" but when I click on it, there's nothing to expand. No reasoning trace, no thinking steps, nothing.

With original OpenAI models (e.g. gpt-5.5), it works normally - For example i see "Worked for 33m 6s" and when I click on it, there's a full expandable section showing all the reasoning/thinking steps.

What I expected

The expandable reasoning trace to appear the same way it does with native OpenAI models, since the README says reasoning tokens are supported.

What I tried

• Different proxy models (DeepSeek v4 flash free, Mimo v2.5 free) - same result on both
• The timing itself works fine ("Worked for 23s" etc), so the connection is working
• Native OpenAI models through the same Codex app work correctly

My setup

• Codex desktop app (latest version)
• Proxy: opencode zen API key through opencodex's hosted proxy
• Models tested: DeepSeek v4 flash free, Mimo v2.5 free

Question

I'm not super experienced with API response formats, so I'm not sure if this is:

1. A bug in how opencodex formats reasoning tokens in the response
2. A limitation of the free-tier models not returning the same reasoning structure as OpenAI
3. Something on Codex's side that doesn't recognize non-OpenAI reasoning token formats
4. Or something else I'm not thinking of

Would love to understand what's going on here. If there's any log or config I can check on my end to help debug, happy to do that. Thanks!

[lidge-jun]

Investigated end-to-end, including the codex-rs consumer side.

Root cause (confirmed): Codex renders the expandable reasoning trace from the Responses reasoning item's summary[] channel. opencodex maps chat-completions reasoning_content to a raw-reasoning item with summary: [] and only content: [{reasoning_text}], so the Codex app shows the "Worked for Xs" timer (driven separately by elapsed time / reasoning_tokens) but has nothing to expand.

I traced this through codex-rs main to be sure it isn't a client bug:

• codex-api/src/sse/responses.rs parses both response.reasoning_summary_text.delta (→ summary) and response.reasoning_text.delta (→ content).
• core/src/session/turn.rs + core/src/event_mapping.rs carry content into raw_content (not dropped), but the persisted/expandable block comes from the summary channel (tui/src/history_cell/messages.rs ReasoningSummaryCell).
• Native OpenAI models populate summary; routed chat models don't, which is exactly the report.

Fix (planned, opencodex-side only): route chat reasoning_content through the summary channel so the final reasoning item has a non-empty summary[] (treat it like the existing thinking_delta path). Non-OpenAI chat providers don't send a separate condensed summary, so their full reasoning_content is the human-visible thinking and belongs in summary. No codex-rs change needed.

Two sub-cases (look identical to the user):

• (A) Provider returns reasoning_content (DeepSeek-R-style) → opencodex formatting gap, fixed by the above.
• (B) Provider returns no reasoning at all (many free models) → nothing to expand; "Worked for Xs" is pure elapsed time. Model limitation, not an opencodex bug.

To confirm which applies to a specific model, a raw upstream chunk showing whether choices[].delta.reasoning_content is ever present would help. Fix for (A) is planned.

[lidge-jun]

@Rezhnn I couldnt see reasoning summary on app or cli I succeed to parse on exec mode could you hand me a some screenshot and config set?