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
description: A patch release adding Google Gemini 3 Pro Image Preview support, Minimax provider support, and various bug fixes.
3
+
keywords:
4
+
- roo code 3.33.3
5
+
- google gemini 3 pro
6
+
- image preview
7
+
- minimax
8
+
- bug fixes
9
+
image: /img/v3.33.3/v3.33.3.png
10
+
---
11
+
12
+
# Roo Code 3.33.3 Release Notes (2025-11-20)
13
+
14
+
This release introduces image preview support for Google Gemini 3 Pro, adds Minimax as a new provider, and includes several important bug fixes and quality of life improvements.
We've added support for image generation with Google Gemini 3 Pro image generation model ([#9440](https://github.com/RooCodeInc/Roo-Code/pull/9440)):
21
+
22
+
-**Visual Verification**: You can now see previews of images generated by Gemini 3 Pro directly within the interface.
23
+
-**Enhanced Workflow**: This improvement streamlines the creative process by allowing for immediate visual feedback on generated assets.
24
+
25
+
This addition ensures a more seamless experience when working with the latest image generation capabilities from Google.
26
+
27
+
## QOL Improvements
28
+
29
+
***Minimax Provider Support**: Added support for Minimax as an Anthropic-compatible provider ([#9455](https://github.com/RooCodeInc/Roo-Code/pull/9455))
30
+
***Reasoning History**: Reasoning is now stored in conversation history for all providers ([#9451](https://github.com/RooCodeInc/Roo-Code/pull/9451))
31
+
***Improved Read File Tool**: Enhanced the `read_file` tool description with clearer examples ([#9422](https://github.com/RooCodeInc/Roo-Code/pull/9422))
32
+
33
+
## Bug Fixes
34
+
35
+
***Environment Details Duplication**: Fixed an issue where `environment_details` were duplicated when resuming cancelled tasks ([#9442](https://github.com/RooCodeInc/Roo-Code/pull/9442))
36
+
***Preserve Reasoning Flag**: Fixed the `preserveReasoning` flag to correctly control API reasoning inclusion ([#9453](https://github.com/RooCodeInc/Roo-Code/pull/9453))
37
+
***OpenAI Parallel Tool Calls**: Prevented OpenAI Native parallel tool calls when using native tool calling to avoid errors ([#9433](https://github.com/RooCodeInc/Roo-Code/pull/9433))
38
+
***Search and Replace Parsing**: Improved symbol parsing in the search and replace functionality ([#9456](https://github.com/RooCodeInc/Roo-Code/pull/9456))
39
+
***Skipped Tool Results**: Fixed an issue where `tool_result` blocks were not sent for skipped tools in the native protocol ([#9457](https://github.com/RooCodeInc/Roo-Code/pull/9457))
40
+
***Markdown Formatting**: Improved markdown formatting and added reasoning support ([#9458](https://github.com/RooCodeInc/Roo-Code/pull/9458))
41
+
42
+
## Misc Improvements
43
+
44
+
***Dependency Update**: Updated `glob` dependency to ^11.1.0 ([#9449](https://github.com/RooCodeInc/Roo-Code/pull/9449))
45
+
***Dependency Update**: Updated `tar-fs` to 3.1.1 via pnpm override ([#9450](https://github.com/RooCodeInc/Roo-Code/pull/9450))
Copy file name to clipboardExpand all lines: docs/update-notes/v3.33.mdx
+11-1Lines changed: 11 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -24,10 +24,14 @@ Roo Code now supports Google’s Gemini 3 Pro Preview model through direct Gemin
24
24
-**Strong eval performance**: Achieves a 100% score on internal Roo Code evals and 76.2% on SWE-bench Verified, giving more consistent solutions on real-world coding tasks.
25
25
-**Reliable tool usage**: Executes complex multi-step tool workflows without getting stuck or losing track, especially in long, tool-heavy tasks.
26
26
-**Better out-of-the-box defaults**: Uses `gemini-2.5-pro` by default where supported, sets a more natural temperature of 1, cleans up the Gemini model list, and includes reasoning / “thought” tokens in cost reporting so usage numbers better match provider billing.
27
+
-**Image Preview Support**: Added support for image previews with Google Gemini 3 Pro image generation models ([#9440](https://github.com/RooCodeInc/Roo-Code/pull/9440)).
27
28
28
29
## QOL Improvements
29
30
30
31
-**Native Tool Calling for OpenAI-compatible Providers** ([#9369](https://github.com/RooCodeInc/Roo-Code/pull/9369)): Enables native tool use with a wider range of models and providers that support the OpenAI format, offering more robust tool interactions in both streaming and non-streaming modes.
32
+
-**Minimax Provider Support** ([#9455](https://github.com/RooCodeInc/Roo-Code/pull/9455)): Added support for Minimax as an Anthropic-compatible provider.
33
+
-**Reasoning History** ([#9451](https://github.com/RooCodeInc/Roo-Code/pull/9451)): Reasoning is now stored in conversation history for all providers.
34
+
-**Improved Read File Tool** ([#9422](https://github.com/RooCodeInc/Roo-Code/pull/9422)): Enhanced the `read_file` tool description with clearer examples.
31
35
32
36
-**Git status in environment details** ([#9310](https://github.com/RooCodeInc/Roo-Code/pull/9310)): Shows git status information in environment details so agents have more context about untracked, modified, and staged files.
33
37
-**Tool protocol selector in advanced settings** ([#9324](https://github.com/RooCodeInc/Roo-Code/pull/9324)): Lets you choose which tool protocol to use (such as XML vs native) without editing config files.
@@ -58,4 +62,10 @@ Roo Code now supports Google’s Gemini 3 Pro Preview model through direct Gemin
58
62
-**Use shields.io badges instead of badgen.net** ([#9318](https://github.com/RooCodeInc/Roo-Code/pull/9318)): Replaces broken badge URLs with shields.io.
59
63
-**Cap git status file sampling in evals** ([#9322](https://github.com/RooCodeInc/Roo-Code/pull/9322)): Adds a maximum for git status files in eval settings.
60
64
-**Preserve images in native tool call results** ([#9401](https://github.com/RooCodeInc/Roo-Code/pull/9401)): Ensures images attached to messages are preserved and passed through native tool calls so follow-up reasoning can still use visual context instead of losing it as plain text only.
61
-
-**Faster router model metadata loading** ([#9410](https://github.com/RooCodeInc/Roo-Code/pull/9410)): Reduces redundant getModel() calls and adds a disk-backed metadata cache so router-based providers start faster and are more reliable when loading model info.
65
+
-**Faster router model metadata loading** ([#9410](https://github.com/RooCodeInc/Roo-Code/pull/9410)): Reduces redundant getModel() calls and adds a disk-backed metadata cache so router-based providers start faster and are more reliable when loading model info.
66
+
-**Environment Details Duplication** ([#9442](https://github.com/RooCodeInc/Roo-Code/pull/9442)): Fixed an issue where `environment_details` were duplicated when resuming cancelled tasks.
67
+
-**Preserve Reasoning Flag** ([#9453](https://github.com/RooCodeInc/Roo-Code/pull/9453)): Fixed the `preserveReasoning` flag to correctly control API reasoning inclusion.
68
+
-**OpenAI Parallel Tool Calls** ([#9433](https://github.com/RooCodeInc/Roo-Code/pull/9433)): Prevented OpenAI Native parallel tool calls when using native tool calling to avoid errors.
69
+
-**Search and Replace Parsing** ([#9456](https://github.com/RooCodeInc/Roo-Code/pull/9456)): Improved symbol parsing in the search and replace functionality.
70
+
-**Skipped Tool Results** ([#9457](https://github.com/RooCodeInc/Roo-Code/pull/9457)): Fixed an issue where `tool_result` blocks were not sent for skipped tools in the native protocol.
71
+
-**Markdown Formatting** ([#9458](https://github.com/RooCodeInc/Roo-Code/pull/9458)): Improved markdown formatting and added reasoning support.
0 commit comments