|
| 1 | +--- |
| 2 | +description: Combined notes for Roo Code 3.32, including GPT-5.1 model support, OpenAI Responses prompt caching, and reliability fixes for OpenAI Native and native tools. |
| 3 | +keywords: |
| 4 | + - roo code 3.32 |
| 5 | + - release notes |
| 6 | + - gpt-5.1 |
| 7 | + - openai prompt caching |
| 8 | + - bug fixes |
| 9 | +image: /img/v3.32.0/v3.32.0.png |
| 10 | +--- |
| 11 | + |
| 12 | +# Roo Code 3.32 Release Notes (2025-11-14) |
| 13 | + |
| 14 | +Roo Code 3.32 combines GPT-5.1 model support and OpenAI Responses prompt caching with reliability fixes for OpenAI Native and native tool protocol. |
| 15 | + |
| 16 | +<img src="/img/v3.32.0/v3.32.0.png" alt="Roo Code v3.32 Release" width="600" /> |
| 17 | + |
| 18 | +## GPT-5.1 models and OpenAI prompt caching |
| 19 | + |
| 20 | +* **GPT-5.1 in Roo Code** ([#9252](https://github.com/RooCodeInc/Roo-Code/pull/9252), [#9258](https://github.com/RooCodeInc/Roo-Code/pull/9258), [#9261](https://github.com/RooCodeInc/Roo-Code/pull/9261), [#9259](https://github.com/RooCodeInc/Roo-Code/pull/9259)) |
| 21 | + Adds the new GPT-5.1 models across supported providers, with extended 24‑hour prompt caching support in the OpenAI Native provider and improved overall software‑engineering performance. |
| 22 | + |
| 23 | +## Bug Fixes |
| 24 | + |
| 25 | +* **Share button works reliably again** ([#9253](https://github.com/RooCodeInc/Roo-Code/pull/9253)) |
| 26 | + Fixes the Share button so you can open the share popover and send shared tasks and messages without failures. |
| 27 | + |
| 28 | +* **Keep OpenAI Native reasoning stable during condensing** ([#9263](https://github.com/RooCodeInc/Roo-Code/pull/9263)) |
| 29 | + Keeps encrypted reasoning blocks paired with the assistant messages that produced them so OpenAI Native condensing no longer throws errors and long-running tasks with reasoning stay stable even as history is trimmed. |
| 30 | + |
| 31 | +* **Prevent duplicate tool_result blocks for read_file in native protocol** ([#9272](https://github.com/RooCodeInc/Roo-Code/pull/9272)) |
| 32 | + Ensures read_file emits exactly one tool_result per tool_call_id in native protocol, preventing 400 errors while still surfacing clear error messages to both the agent and the user. |
| 33 | + |
| 34 | +* **Fix duplicate tool blocks causing "tool has already been used" errors** ([#9275](https://github.com/RooCodeInc/Roo-Code/pull/9275)) |
| 35 | + Stops XML protocol runs from merging duplicate tool blocks into assistant messages so each tool block executes only once per turn and tool-driven tasks no longer get stuck behind false “tool has already been used” errors. |
| 36 | + |
| 37 | +* **Make OpenAI Native cancellation stop streaming immediately** ([#9276](https://github.com/RooCodeInc/Roo-Code/pull/9276)) |
| 38 | + Adds an abort controller so cancelling an OpenAI Native request stops streaming promptly, avoiding extra tokens after cancellation and reducing wasted usage. |
| 39 | + |
| 40 | +* **Disable XML parser for native tool protocol** ([#9277](https://github.com/RooCodeInc/Roo-Code/pull/9277)) |
| 41 | + Bypasses XML parsing when toolProtocol is set to native and relies on structured tool_call chunks instead, avoiding mis-parsed XML-like snippets and making native tool runs more reliable—especially for models that sometimes mix formats. |
0 commit comments