Hey rajeshb89,
The error "DuckDB file not created in OPFS after all init operations." should be fixed. Please update and try again. I recently rewrote the vector db implementation completely which causes several bugs. Sorry, and thank you for the report!

Hey User #219606,
Anthropic is not tested. However, it should still work with the plugin. There is code to support claude.

Hey Gustavo Carnevali,
The issue should be fixed!

Hey, would it be possible to add support for the Requesty LLM provider? It is similar to Openrouter, even has Open AI like chat completions API: https://docs.requesty.ai/applications/openwebui

Hi,
I get the following pop-up when I click on Sync LocalVecDB - OPFS is not persisted. The browser may delete the indexed data. For better reliability, it's recommended to enable persistent storage. Do you want to ignore this and continue anyway?

When I click Continue, the sync completes but then I have DB Status as Error - When I type something in search I get, Error: Error: DuckDB file not created in OPFS after all init operations.

I have tried both Google AI and Pinecone for the Embedding. I have tried Edge as well as Chrome. This is in a Windows machine.

Any idea what the issue could be?

Does this work with Anthropic's Claude.ai? If not, would it be too much trouble to make it work? Pleeeeease :D

Hello, and thank you for developing this plugin.

I followed all the instructions for synchronizing my notes to LocalVecDB (using my embedding API URL and key) on the browser version installed as a Google Chrome app on my MacBook Pro; however, I have a rather extensive library of notes (?) (4,558), and the sync freezes and will not go past 2,047 notes. Once it reaches this point, the app does not respond, and I must close it.

I tried the sync operation on the Mac version of the app, and it also failed.

Any thoughts on how I can sync my entire notes library? Thank you.

Hey kasimasi2,

Sorry you're running into issues, and thanks for bringing it up!

> Error: Invalid URL (POST /v1/chat/completions/chat/completions)
This usually means the OpenAI API URL is set incorrectly. Please make sure the OpenAI LLM API URL in your settings is exactly:

https://api.openai.com/v1

Don't include /chat/completions in the URL.

> How exactly do I get a Google AI Studio embedding API key? I
Here's a short video guide that walks through the process: https://youtu.be/3J_4xPv-KPg?si=539mi3e-A2oJCDKr

I tried using this plugin a few months ago, and again just now, but, I keep getting errors, even after syncing LocalVecDB. Running an @notes query just gives me this error:

Error: Invalid URL (POST /v1/chat/completions/chat/completions)

I have my OpenAI key set up, but it still doesn't work.

How exactly do I get a Google AI Studio embedding API key? I went to their website and I tried generating a Gemini API key, but how can I tell if this is the right service/type of API key that I'm using? Google has a LOT of different websites/products, so it's very confusing...

> Could you extend the condition to use the OpenAI client to also apply for Open Router URLs? It works perfectly, just needs the check to not throw the error, I'd prefer to not use a modified version.

This has been added. You need to provide https://openrouter.ai/api/v1 in LLM API URL to use this.

> Is there any way to make the chat take 100% of the peek viewer panel's height?

Not at the moment. This is a restriction in Amplenote Plugin API. You can create a feature request for this to amplenote team.

Hi!

I'll start saying that I've been using the plugin for the last couple of weeks and love it!

Quick question: Is there any way to make the chat take 100% of the peek viewer panel's height?

Thanks!

Amazing plugin!

Could you extend the condition to use the OpenAI client to also apply for Open Router URLs? It works perfectly, just needs the check to not throw the error, I'd prefer to not use a modified version.

Thank you for the quick response! I was just coming back to update since I found it in the plugin doc...someone didnt RTFM lol.

Long shot but since you answered :), I'm using the app version on a mac and now getting "Failed to sync notes: Error: Error: The device (webgpu) does not support fp16." when I try to sync.

Trying to avoid using pinecone at the moment if possible.

*** In case anyone runs into this ***

I opened Amplenote in the browser and syncing worked fine.

This plugin is great!!!

Hey User #144003,

In order to search your notes, you need to sync the notes to LocalVecDB. This needs to be done periodically.

Check https://public.amplenote.com/jKhhLtHMaSDGM8ooY4R9MiYi#Q)_How_to_sync_to_LocalVecDB%3F and https://public.amplenote.com/16oi13jtaNMoSxqjQMKgBdUE for more details.

I am trying to search my notes using @notes but getting a sync error

LocalVecDB search failed: Error: Error: No syncing has been performed, or the last sync is outdated. Please sync your notes with LocalVecDB.

I'm using the OpenAI model if that makes a difference at all

Is there any chance I could get some info on that?

Incredible plugin! This transforms Amplenote into mem.ai.