ammaarreshi/gemma-chat

Me to. I just give up and deinstalled it.
Where are the the model files stored? I want to clear this completely!

same error on macbook pro m3
OS 26.4.1
node 22.22.1 or 20.20.2

This seems to be a very common error. Just downloaded and installed on M2 MacBook Pro and also seeing this error message when I press the "Download" button in the application. It just fails to download and opens a non-functional app window. Unable to sort out a solution so far!

Error Output

```
[75329:0505/012056.860026:ERROR:CONSOLE(1)] "Request Autofill.enable failed. {"code":-32601,"message":"'Autofill.enable' wasn't found"}", source: devtools://devtools/bundled/core/protocol_client/protocol_client.js (1)
[75329:0505/012056.860069:ERROR:CONSOLE(1)] "Request Autofill.setAddresses failed. {"code":-32601,"message":"'Autofill.setAddresses' wasn't found"}", source: devtools://devtools/bundled/core/protocol_client/protocol_client.js (1)
[mlx] Found mlx-lm in venv
[mlx] Starting server: /Users/thinkyhead/Library/Application Support/gemma-chat/mlx/venv/bin/python3 -m mlx_lm server --model mlx-community/gemma-4-e4b-it-4bit --port 11434
[mlx] Server is healthy
[7...

Not working for me either

> > I put together a Windows preview build that runs Gemma Chat through Ollama instead of MLX:
> > ```
> > * PR: [feat: add Windows support via Ollama runtime #14](https://github.com/ammaarreshi/gemma-chat/pull/14)
> >
> > * Windows installer release: https://github.com/daveroyal/gemma-chat/releases/tag/windows-ollama-v0.1.0
> > ```
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > For Windows users: install Ollama, run `ollama pull gemma4:31b`, then install the unsigned preview `.exe`. SmartScreen may warn because this preview build is not code-signed yet.
>
> How can i change the port that it will listen? For me is trying to get 11434, but this port is not availabe and i couldnt change it

I just updated it so you can edit `GemmaChat\src\main\ollama.ts` and change the port (see line 4 in the file).

https://github.com/daveroyal/gemma-chat/releases/tag/windows-ollama-v0.1.1

> I put together a Windows preview build that runs Gemma Chat through Ollama instead of MLX:
>
> * PR: [feat: add Windows support via Ollama runtime #14](https://github.com/ammaarreshi/gemma-chat/pull/14)
>
> * Windows installer release: https://github.com/daveroyal/gemma-chat/releases/tag/windows-ollama-v0.1.0
>
>
> For Windows users: install Ollama, run `ollama pull gemma4:31b`, then install the unsigned preview `.exe`. SmartScreen may warn because this preview build is not code-signed yet.

How can i change the port that it will listen? For me is trying to get 11434, but this port is not availabe and i couldnt change it

same error for me ? :-(

npm install
npm warn deprecated inflight@1.0.6: This module is not supported, and leaks memory. Do not use it. Check out lru-cache if you want a good and tested way to coalesce async requests by a key value, which is much more comprehensive and powerful.
npm warn deprecated rimraf@3.0.2: Rimraf versions prior to v4 are no longer supported
npm warn deprecated glob@7.2.3: Old versions of glob are not supported, and contain widely publicized security vulnerabilities, which have been fixed in the current version. Please update. Support for old versions may be purchased (at exorbitant rates) by contacting i@izs.me
npm warn deprecated glob@8.1.0: Old versions of glob are not supported, and contain widely publicized security vulnerabilities, which have been fixed in the current version. Please update. Support for old versions may be purchased (at exorbitant rates) by contacting i@izs.me
npm warn deprecated tar@6.2.1: Old versions of tar are not supported, and contain ...

I put together a Windows preview build that runs Gemma Chat through Ollama instead of MLX:

- PR: https://github.com/ammaarreshi/gemma-chat/pull/14
- Windows installer release: https://github.com/daveroyal/gemma-chat/releases/tag/windows-ollama-v0.1.0

For Windows users: install Ollama, run `ollama pull gemma4:31b`, then install the unsigned preview `.exe`. SmartScreen may warn because this preview build is not code-signed yet.

Tested PR #9 (stacked on #8) locally — works great.

- The port-collision fix lets MLX and Ollama coexist (Ollama stayed on 11434, MLX picked the next free port). `npm run dev` now starts cleanly even with Ollama already running.
- Nice side benefit: the Ollama integration auto-discovers models I'd already pulled, so I didn't have to re-download anything to try them in Gemma Chat.

Thanks for the quick turnaround.

---

### Quick security review of PR #9 (informal — sharing in case it's useful)

**Scope reviewed:** `+447/-53` across 9 files, including the new `src/main/ollama.ts`. No new npm dependencies, no new IPC channels.

**No high-risk issues found.** Specifically:
- All listeners stay bound to `127.0.0.1` — no new inbound surface.
- Pure HTTP integration with Ollama; no `spawn`/`exec` introduced.
- `model` flows through JSON bodies, never interpolated into URL paths.
- SSE parsing is defensive (`JSON.parse` wrapped in try/catch, malformed events skipped).
- Removed code (`MLX_UR...

I opened #8 as a focused fix for this: MLX now selects a free local port instead of hardcoding 11434, so it can coexist with Ollama.

I also opened #9 as a separate draft for broader Ollama runtime support, but #8 is the small bugfix PR intended to close this issue.