[GH-ISSUE #18521] issue: Code Interpreter does not function as expected with Gemini pipeline function

Originally created by @swiftdev29 on GitHub (Oct 22, 2025). Original GitHub issue: https://github.com/open-webui/open-webui/issues/18521 ### Check Existing Issues - [x] I have searched for any existing and/or related issues. - [x] I have searched for any existing and/or related discussions. - [x] I am using the latest version of Open WebUI. ### Installation Method Docker ### Open WebUI Version v0.6.34 ### Ollama Version (if applicable) _No response_ ### Operating System macOS ### Browser (if applicable) Safari ### Confirmation - [x] I have read and followed all instructions in `README.md`. - [x] I am using the latest version of **both** Open WebUI and Ollama. - [x] I have included the browser console logs. - [x] I have included the Docker container logs. - [x] I have **provided every relevant configuration, setting, and environment variable used in my setup.** - [x] I have clearly **listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup** (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc). - [x] I have documented **step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation**. My steps: - Start with the initial platform/version/OS and dependencies used, - Specify exact install/launch/configure commands, - List URLs visited, user input (incl. example values/emails/passwords if needed), - Describe all options and toggles enabled or changed, - Include any files or environmental changes, - Identify the expected and actual result at each stage, - Ensure any reasonably skilled user can follow and hit the same issue. ### Expected Behavior Code Interpreter should load up with proper tags, analyzing block and render output properly. ### Actual Behavior Tested with and without the Gemini Pipeline function. First, I tried using the Gemini's OpenAI Compatible API. The same model (gemini-2.5-flash) worked fine with the Code Interpreter, except with rendering results like images. However, with the Gemini Pipeline function (https://openwebui.com/f/owndev/google_gemini), the appropriate tags for code interpreter appear as it is in the model response. No code is run. <img width="490" height="510" alt="Image" src="https://github.com/user-attachments/assets/f8ffe355-a192-4f2d-9dee-b14b89fabce2" /> ### Steps to Reproduce 1. Use the Gemini Pipeline Function (https://openwebui.com/f/owndev/google_gemini) 2. Setup with API Key 3. Enable Code interpreter and input a question to the model. ### Logs & Screenshots <img width="1129" height="486" alt="Image" src="https://github.com/user-attachments/assets/ee977c4a-96ef-49db-bc4b-b4f2592daabd" /> <img width="490" height="510" alt="Image" src="https://github.com/user-attachments/assets/61728df6-31e6-43f8-ac1a-d4ecd23ede2c" /> ### Additional Information _No response_

GiteaMirror commented

2026-04-25 08:04:20 -05:00

Owner

Originally created by @swiftdev29 on GitHub (Oct 22, 2025).
Original GitHub issue: https://github.com/open-webui/open-webui/issues/18521

Check Existing Issues

I have searched for any existing and/or related issues.
I have searched for any existing and/or related discussions.
I am using the latest version of Open WebUI.

Installation Method

Docker

Open WebUI Version

v0.6.34

Ollama Version (if applicable)

No response

Operating System

macOS

Browser (if applicable)

Safari

Confirmation

I have read and followed all instructions in README.md.
I am using the latest version of both Open WebUI and Ollama.
I have included the browser console logs.
I have included the Docker container logs.
I have provided every relevant configuration, setting, and environment variable used in my setup.
I have clearly listed every relevant configuration, custom setting, environment variable, and command-line option that influences my setup (such as Docker Compose overrides, .env values, browser settings, authentication configurations, etc).
I have documented step-by-step reproduction instructions that are precise, sequential, and leave nothing to interpretation. My steps:
Start with the initial platform/version/OS and dependencies used,
Specify exact install/launch/configure commands,
List URLs visited, user input (incl. example values/emails/passwords if needed),
Describe all options and toggles enabled or changed,
Include any files or environmental changes,
Identify the expected and actual result at each stage,
Ensure any reasonably skilled user can follow and hit the same issue.

Expected Behavior

Code Interpreter should load up with proper tags, analyzing block and render output properly.

Actual Behavior

Tested with and without the Gemini Pipeline function. First, I tried using the Gemini's OpenAI Compatible API. The same model (gemini-2.5-flash) worked fine with the Code Interpreter, except with rendering results like images.

However, with the Gemini Pipeline function (https://openwebui.com/f/owndev/google_gemini), the appropriate tags for code interpreter appear as it is in the model response. No code is run.